I just thought I'd give my thoughts on it especially since numbers are down. Just from checking the standings this is how many teams I think we've had

TL76 = 21 (+1 late entry)
TL75 = 24
TL74 = 28
TL73 = 30
TL72 = 30
TL71 = 31

As a result of this we've had sections narrowing. One other problem is we've had a few teams drop out at the top and bottom of the rankings. This has actually led to an arms race to avoid certain sections.

For example I try to avoid having two teams in the Fischer Section. In my case main motivation is partly that I can't field 2 strong teams with winning chances and also that some players ask to play for two teams and you obviously want them in first and second section. Looking in Fischer Section I think only first 4 teams actually aimed to be in it. The rest are either teams who can't manipulate their rating (i.e. fixed player pool) or those caught out who thought they shed enough points and realised they didn't. For example in TL75 I assumed anything under 2k should be fine to avoid it and was wrong. In TL76 I figured 1950ish would be OK and went in at 1964 and got burnt. In TL77 I will plan to aim at around 1900ish to be safe. The only problem and where the arms race comes into it is that the teams who have ratings 7th, 8th, 9th and 10th in average rating happen to be 2nd/3rd teams of teams who are 1st, 2nd and 3rd in average rating and in Fischer Section. If you want more fun the 11th team also has overlap with 4th team.

This basically means that all the teams at the top of the Spassky section are desperately trying to avoid getting into the Fischer Section. When I lower average rating of my teams, everyone else counters and eventually we end up with extreme gaps and dead rating zones where no team wants to be in. Team who gets the short stick is then forced to sacrifice a team like Mysterious Queen Moves as it's better to cripple one team than to cripple both of them when you have a player clash.

This also applies at the bottom of TL. It used to be that I could aim for a team to be close to 1700 and it would be quite competitive. To give an example in TL71 my bottom team had an average of 1693 and this season my lowest team has a rating of 1765 and despite that is actually 3rd lowest in section and almost 100 points off the pace of top team in section.

In this case my solution is quite clear, next season my lowest team needs an average of around 1800.

However taking it to the logical extreme as my "Spassky" team has had to lower itself from U2000 to U1950 to U1900 and my lowest team has had to increase from 1650 to 1700 to 1750 to 1800 aren't they soon going to meet in the middle?

Then I guess next question is that if all teams will be aiming for 1800 to 1900 sweet spot would fixed sections make sense?

I'm a bit interested in how other captains see it and if they have any alternative thoughts or approaches to how I see it.

Yes, the number of teams have been decreasing so far, and, moreover, TL76 is the first season comprising just 3 sections. Such situation is particularly unfavorable for big teams, consisting of 3 and more squads. It's mathematically impossible not to have your squads clashed if 4 of them compete in 3 sections. Even our team who always have 3 squads faces a risk of unfavorable clash and it's just a miracle (and hard work!) that still we were able to sort it out.
As there's a big probability of having just 3 sections in future there's no good solution for a team comprising more than 3 squads, at all. And this is sad.

The clue is, as KRMCHESS mentioned, narrowing of Fischer Section, in particular, so that this season only first 4 teams actually aimed to be in it. In the past a solution was discussed about sections with less number of teams (that time it was related to the lowest-rated section) which compete in a double Round Robin tourney. Perhaps, the same solution could be considered in relation to Fischer, though at the same time there are disadvantages here as well (not a big fun in facing the same opponent two times).

Still, such arrangement with 4 sections (small Fischer with double RR tourney + 3 lower sections with RR tourney) could help to distribute teams in a more accurate manner, IMHO.

Even in the past it was not rare to see a team designed for Spassky finding itself in Fischer. For instance, in T67, when 31 teams competed, the Fischer section went all the way down to 1975, producing a similar situation to the present. On the other hand, in T71 we had a sort of ideal scenario for Fischer: 7 teams above 2150.

Now, as you noted, KRM, many top teams are missing (FlamingPhoenixFighters, Marines, CafeMasters, BlindStars, JokerSlam) and a peculiar scneario emerged: 4 teams above 2100 (the ones indeed willing to be in Fischer), a vacuum in the 2000-2100 range, and all teams willing to be in Spassky under 2000. Then it's worth mentioning that the total number of teams plays a key role in how they're distributed, and with 21 I see there was no other reasonable solution for admins aside from 3 x 7. If we had 22 right away, 4-6-6-6 would be a possible scheme. Sections with 4 teams lack a bit of variety and won't see a playoff, and sections with 6 give everybody fewer games, but they provide more balance than sections with 7 or 8 teams, which are attractive regarding playing opportunities. I guess one can't have everything...

I never played under the old system and therefore don't know how well it worked, but, if we don't want a big gap among teams, I think sections separated by 100 points instead of 200 would make more sense, e.g., Open, U2000, U1900, U1800, U1700, or Open, U2050, U1950 etc. I don't know, maybe this could encourage teams that currently have only one group playing to go with two? If you have a number of players willing to be in two sections that could work, as there would be no fear of an udesirable clash, and as a result the overall number could become considerably higher. But it's just a thought, really.

I think more communication among captains here before seasons would also be beneficial. No need for anyone to go into detail about their team, but we could share if we're going to play or not, if we have 2 or more groups etc. Now that there are fewer of us (in TL and FICS), it's good to work towards a sense of community and keep it going. And let me share good news against the dropping numbers: Singularity will most likely be back next season, and maybe EstrellasLatinAmerica will have a second group enlisted.

Number of teams in recent seasons: T71- 30; T72- 28; T73- 30; T74- 29; T75- 24.

Best,
Beto.

Well to be honest I don't think there are enough teams to have too many sections. Considering that in current TL system we have 3 to 4 sections it means that realistically we should aim to have 3 to 4 rating cut offs. If going for 4 sections I would suggest Open, U2000, U1900 and something like U1750 to 1800. A 4 Team Fischer double RR does sound more interesting and there is a precedent with Polgar Section.

Communication between captains could be useful but as I mentioned in first post until we reach 12th team in Spassky section any team from 7th to 11th added to Fischer section would have actually clashed with a Fischer team and if the 7th to 11th ranked teams conspired to allow TrickyMove with an average rating of 1874.25 to enter the Fischer Section I doubt they would be happy! I think some team's decisions not to enter were very last minute as I know one person who joined my team this season only found out he wasn't in a team when sections went up.

I'll also note that having two teams in the same section isn't a major problem for me. However what is a problem is having two teams unexpectedly in the same section. For example in TL75 I planned to have 2 teams in lowest section and 1 in each of the other 3. As two teams in lower section were planned in advance I had no clashes and worked well. Unfortunately my Spassky team ended up in Fischer Section. This meant I lost a couple of players due to clashes in the top section. It also meant I could no longer get people from my Spassky team to cover as I normally do and I also couldn't get people in Kasparov section to cover as they had to be drafted into Spassky team who had lost players due to clashes and thus were ineligible for my top Fischer team. This meant I had to go down to my 4th and 5th team for cover as a sub and in the end I had to draft in a couple of 1600s to avoid defaults.

That's also the big reason I'm a fan of fixed ratings. It means that if a player plays in 2 teams as long as you get them on both sides of a fixed rating you know you'll never get a clash. For example Mysterious_Queen_Moves due to player clashes lost their top board and now has a rating 1912.75 in Fischer Section while Spassky Section has a 1964.00 max. As a captain I actually feel quite guilty fielding a team in a section fully aware they are 300 points average lower than top team when other team is 2221.25 and favourite to win the section. However if I averaged them out then I get 2 teams that are 2067 average and neither is likely to win.

If I was ultra-competitive what I would do is weaken Mysterious_Queen_Moves even further and drop players into Mysterious_Rook_Moves (swapping thatgirl and Miltie takes team up to 1960) and Mysterious_Pawn_Moves (swapping PankracyRozumek and Miltie takes team up to 1846.25) to take both teams right up to the rating limit and drop Mysterious_Queen_Moves to the 1700s. Note that while legally permitted, it is a bit immoral so I don't plan to do so.

As you can see above with unexpected clashes in TL75 I tried to get both teams to work and in the end neither worked with forfeits in 1 team when a new player went AWOL and having to get 1600s in other team. In TL76 having learnt from experience I've instead sacrificed a team to make sure other works OK. Neither is really an enviable choice. I'm planning to heavily reduce ratings of teams I submit next season so I'll avoid problem in TL77 but I think arithmetic means that if I avoid it someone else gets it in my place. That's why it might be better to get a solution that harms nobody rather than a selfish solution that keeps me alone out of harms way.

Yes, KRM, I remember you had two teams in Alekhine in T75, and it was the first time I saw Mysterious_Knight_Moves in action. :) It's the same strategy TheRangers have been adopting recently, and of course when squads of the same team are planned to be together there's no harm. In a scenario in which so many teams disappeared, it's great that you and your mates can still form 4-5 teams, and we certainly don't want you to get discouraged to submit more teams because of these undesirable clashes.

In fact, the situation in the current season could be worse: if our team NewBlood/Erythro had the same average as in previous seasons, Mysterious_Rook_Moves would end up as the top team in Alekhine (as Szachowe.pl was not in when sections were created), so you'd have all your teams in the top and bottom sections, none in the central one. For us it'd be very bad too because, as you can see, araszek and oje are in both NewBlood/Leuko and NewBlood/Erythro. We were lucky because in the process of planning and changing Leuko for the new season Erythro's average dropped and neither PreZandy nor I saw any problem in having it in Alekhine instead of Kasparov, not predicting there would be only 3 sections this time.

So, the more I think about it, the more I like the idea of sections with fixed rating caps. My doubt is: what happened in the past (T44-45) that led to a change in the system?

I cannot find the relevant forum thread, but as far as I can remember the problem was that some team captains (especially team captains of larger teams) were gaming the system by aiming to submit teams with average ratings of 1990-1999, 1790-1799 etc. Small teams were at disadvantage, and Mysterious Moves, which I led then, was a small team :), so I was huge supporter of the new system. I didn't expect that this new system would led to a different set of problems.

I don't like the current situation a lot, but I don't think that moving back to fixed rating caps is any better or worse than the current system. I would rather try one of these:

1. make it a team Swiss tournament with only one section, just like it is on lichess4545. They have around 20-25 teams competing in 8 round Swiss tournament. Disadvantage: it would be hard to maintain the current week extensions. Another disadvantage: people wouldn't be allowed to play in more than one team. Yet another disadvantage: probably software changes would be necessary in TL bot and website.

2. keep it a round robin tournament (it shouldn't matter if we use fixed caps or not), but change the way the teams are drafted: first make players sign up (individually, not as teams), then select team captains (there are multiple ways to do so, we would figure it out), finally the captains would draft non-captains to form teams. Alternatively the non-captains could be added to captains automatically, to minimize the team rating differences. The disadvantage for some would be that it would not be possible to keep the teams stable between seasons. But I wouldn't personally mind to try that. A nice thing is that if there is a consensus that we have a problem, we can try to sort out this problem without adding extra burden to the TL staff.

The Fischer section might look like that, assuming that we kept the same captains, and all captains would selfishly choose the best available player, and that captains would choose player for specific board in order of increasing rating of already existing boards:
schachbjm (2405) - PeterTrsavec (2138) -Tinours (2031) - Benneyboy (1978)
pchesso (1873) - xandor (2458) - PeterSanderson (2064) -Slek (1990)
kurumim (1969) - oldengawy (2218) - KRMCHESS (2081) - damouno (2021)
smallblackcat (2237) - Maras (2162) - MichaelOgbeide (2056) - Spudweb (1987)
milpat (1915) - axeltiger (2238) - LouisXIV (2117) - Shakazahn (2019)
blore (1947) - jaberwock (2234) - Yaro (2104) - Misteraw (2014)
PankracyRozumek (2069) - Iwulu (2185) - Oakwell (2071) - Ymar (2003)

The rating averages would be: 2138, 2096, 2072, 2110, 2072, 2074, 2082.

Cheers,
Michał

Well I'll give a few thoughts. First of all for the completely open section it does heavily favour the higher rated (Fischer) teams. One thing that could be considered is that for example in one competition for my otb team we have a Summer Cup and way points are allocated is that if a player is over 100 points lower rated they an extra 1/4 point and if they are over 200 points lower rated then they start with an extra 3/4. That does mean that if a team outranks all 4 opponents by 200+ points it starts at 3-0 for lower rated team so a 3.5-0.5 result ends up as 3.5-3.5 draw and does in theory mean that teams aren't penalised too heavily if they have weaker players in them.

As far as the draft proposal goes it is interesting as a concept. Only drawback is that player's availability might differ as you could get some strong players who can't commit to playing every week. As far as how draft would occur I think that unless you can get all the captains to commit to appearing at a certain time only way captains could choose would be to rank all the players in their preferred order so automatic assignment might be better with captains then placed in appropriate teams. There is also problem that if captains play in the team then if strength of captains varies that could influence teams as you have several captains in 1500 to 1600 range and at other extreme have a few who are 2200+.

Actually when thinking about it another benefit of the draft system you've suggested is that new players have a much lower entry threshold to get into TL. Currently you need to message several captains of TL teams and hope one of them has space. If it's a draft approach then a new player can sign up and they'll be assigned a team. Even if we revert to old system afterwards it is possible that draft could get new players into teams since once you've played a season in draft and you're a free agent it would mean teams would be more likely to proactively recruit them, especially if they're already teammates.

I don't want to rule out, or rule in any of these ideas yet, but I'll just say that returning to fixed rating sections has one definite advantage from my perspective - it wouldn't require changing anything. People could submit teams just as they do now, and administratively everything would work the same way.

I cannot find the relevant forum thread, but as far as I can remember the problem was that some team captains (especially team captains of larger teams) were gaming the system by aiming to submit teams with average ratings of 1990-1999, 1790-1799 etc

This is true, but at the time we were also quite concerned about the viability of the top and bottom sections (sometimes the Open (+2000) section wouldn't even get 4 teams, and we frequently didn't run U1400 at all due to lack of interest). The middle sections by contrast got very bloated (I think we had 17 teams in U2000 once).

Now we have a different set of problems. For one thing, our lowest rated teams now would not even fit under the second lowest rating cap (U1600) under the old system. It seems to me that one way we could improve numbers is at the lower end of the rating scale. Also, I don't think we have any fear of a 17-team section anymore.

There are still drawbacks to be considered. Smaller teams tend to be worse off with fixed sections, as Michal notes. There's less control over sections sizes (if 3 teams or 5 teams enter a certain section, we can't really do anything about it). Recruiting would play more of a role, as captains had incentives to 'game the system', as Michal noted. Anyway we have a good stretch of time to consider these matters, as I don't anticipate opening T77 until February.

Michal and sbc, thanks for the feedback on the old system. As I mentioned, sections used to be separated by 200 points, but, if we separate them by 100 points (U2000, U1900 etc.), I don't think ratings would be a problem. For example, a team with 1990 in average is hardly an outright winner against one with 1930, plus it's not like we have smaller gaps in the current system.

The real issue with the old system is the one pointed out by sbc: there's no control over the number of teams in each section, and more often than not we can run into numbers that are totally unfavourable to our 7-round scheme. I see that in the past there would be divisions within sections, but it's another reality now and I don't have an answer to how sections with 3, 5 or 9 teams could be handled.

In continuing with the non-fixed section caps, I'd suggest (agreeing with PreZandy and KRM) that a Fischer section with just 4 teams should be strongly considered in the future, especially if this vacuum in the 2000-2100 range persists.

Michal, regarding the system you described, I for one really like the identity and history teams have here, so I'm not keen on random assortments of players.

KRM: I like the idea of compensating lower-rated teams with handicap points. It would help at least in non-Fischer sections. In Fischer section, I am still not sure if I would appreciate the current situation of Mysterious Queen Moves :).

sbc: yes, that's it - odd sizes of sections, thanks for bringing it up.

Kurumim, I certainly understand your reasons.

Old System, Fixed Caps:
Pros:
1. Relatively even match-ups as most teams try to get somewhere near the top of the desired ratings cap. More balanced games.
2. Easier to plan for super-teams. Playing in multiple divisions is possible without worrying about playing yourself or having one player on two teams.
3. Playoffs make more sense if you have 2+ divisions of UXXXX (rather than a top-2 rematch).
Cons:
1. Small teams at a disadvantage for "gaming" the ratings (but that seems to be the case now as the number of teams shrink?)
2. Unequal sections. High and low sections may have very few teams. Many teams in the middle.

Current System:
Pros:
1. In theory easy to submit because there's no reason to game the system when you don't know the divisions.
2. Balanced number of teams between sections. Rather than have a two round playoffs for the four winners of four different UXXXX section... and then having the lonely 4-team Open section trying to figure out how to make their playoffs relevant after they've already played everyone twice and had a bye...
Cons:
1. In reality, just as difficult to submit for super-teams that want to avoid playing themselves or just want to allow players to play on more than one team.
2. Playoffs are just a rematch of the best teams.
3. Does not seem to encourage new players like I thought it might, as there is less incentive to recruit to meet certain rating requirements.
4. Potentially uneven team/individual match-ups.

lichess system:
Pros:
1. Easier to balance teams and sections.
2. Anyone could sign up; potentially greater recruitment? I personally doubt it, though. I think FICS is just shrinking in general. This forum is an example of that... with this topic being the first post in months.
Cons:
1. Lose the camaraderie of playing for Your Team.
2. Swiss would lose the ability for extensions, perhaps necessitating more players per team, and thus maybe even fewer teams.
3. At that point you might just consider joining the lichess TL45-45 league instead... though I still prefer babaschess and watchbot over any web-based client.

Some kind of UXXXX system where players are auto-drafted a la lichess, but then play according to FICS TL rules under the old system could also be possible, again potentially losing camaraderie.

That said, most players here have been here for years... know and respect the players on other teams... and probably would enjoy playing for their random houses after being subjected to the sorting hat. So maybe it's not so much of an issue. I personally wonder if I would have still found the time to play TL if MFO were still around, or if my other excuses (work mostly, but also my free weekend time often spent playing bridge these days) would have kept me out of it still.

As I mentioned, sections used to be separated by 200 points, but, if we separate them by 100 points (U2000, U1900 etc.), I don't think ratings would be a problem. For example, a team with 1990 in average is hardly an outright winner against one with 1930, plus it's not like we have smaller gaps in the current system.

There's no reason why we couldn't do this. One possible issue is that we simply won't have enough teams to make such narrow sections viable. It's hard to predict how team formation would work under fixed sections, since we haven't had them for so long. However, in the old days virtually all teams would end up within 100 points of the rating cap. We didn't get a lot of sub-1900 teams in U2000 for example, and those teams at least knew what they signed up for.

2. Easier to plan for super-teams. Playing in multiple divisions is possible without worrying about playing yourself or having one player on two teams.
3. Playoffs make more sense if you have 2+ divisions of UXXXX (rather than a top-2 rematch).

The point about super-teams is pretty important. At the time of the change, people like Michal and I were concerned about the disincentives for smaller teams. However, now we have unintended disincentives for super-teams, and since they are responsible for such a large proportion of our numbers, this is really counter-productive.

Also I concur with Joshua, playoffs were more fun under the old system. Tiebreakers presented some issues, but they only became really controversial when semi-finals were needed, and we had to have a decisive and arbitrary tiebreaker like board removal. Finals were either decided by rematch, or we just had shared titles in case of ties.

Some kind of UXXXX system where players are auto-drafted a la lichess, but then play according to FICS TL rules under the old system could also be possible, again potentially losing camaraderie.

Losing camaraderie is a serious concern to me here. This is really what distinguishes TL (and FICS in general) from leagues on other sites. There are also practical issues with the draft model - if automated, it needs coding; if done in person, it's hard to find a time suitable for all captains.

How about the following:

One signup per player
Up to six (eight?) players per roster
Four players per actual weekly lineup
Three tournaments per year
Single Round Robin, extensions
Rating cap U2050
After the deadline, create three sections

Possible effect
Pre-deadline: No need to game the system. A 1950 team won't know what section it will play, but it will be playing competitively in any.
Post-deadline: No arms races, all players stay on the one team they signed up to. New players may still be added throughout the tournament.

Incorporate Snail Bucket
Offer two individual tournaments per year for players who never need a break.

Btw thanks Michał for pointing us to lichess' league. Always good to learn from the others. According to their documentation they have up to 8 players per roster, and also up to 8 games per match, depending on how many players are available on both teams. Plus, they offer players to play a board up or even down.

I don't know how it works in practice, but it sounds clever. Why restrict ourselves to 4 games per match, if more players on both teams are eager to play? We could work out the details and caveats. We could also allow one joker per team to play on other boards, that is 2 games per match, if feasible.

Well I think main problem with fixed cap sections is potential for different sized sections. However here is way I would see us deal with it.

8 Teams = Standard (7 Rounds) + Final
7 Teams = Standard (7 Rounds) inc 1 Bye per team + Final
6 Teams = Standard (5 Rounds) + Semi Final for top 4 (1vs4, 2vs3) then final
5 Teams = Double Round Robin (10 Rounds inc 2 Byes)
4 Teams = Double Round Robin (6 Rounds) + Final

As an interesting thought if you have teams A, B, C, D & E in a 5 team group. Pairings would be as follows

Round 1: E vs C, A vs B, D Bye
Round 2: C vs A, D vs E, B Bye
Round 3: A vs D, B vs C, E Bye
Round 4: B vs E, C vs D, A Bye
Round 5: D vs B, E vs A, C Bye

Now here is interesting part. If you look at round 5 you'll see pairings of D vs B and E vs A. Now team D has a bye in round 1 and team B has a bye in round 2. If you scheduled D vs B in round 1 but gave them 2 weeks to play the game then players from Team D & B can try to schedule games during their bye. In same way for E vs A team E has a bye in round 3 and team A has a bye in round 4. So we could in theory schedule B vs C in round 3 and give both teams 2 weeks to play the game. Then Team C loses their bye but has games every week. This means that it should in theory be possible to compress a 5 team section double round robin into 8 rounds rather than 10 rounds. This means we would end up with:

Round 1: E vs C, A vs B, D vs B (To be played over 2 rounds)
Round 2: C vs A, D vs E
Round 3: A vs D, B vs C, E vs A (To be played over 2 rounds)
Round 4: B vs E, C vs D

In the interest of fairness it's probably best for Team D to have to play a game over 2 rounds in second round robin but I think this could be a viable solution to scheduling. I can also imagine that both sets of players will want game during their bye weekend so there could be a bit of haggling on who yields unless it's played midweek between the two weekends.

Again, one of the main problems that caused this discussion is narrowing of Fischer section. As kurumim noted, the following teams have gone (still we hope some of them will be back one day): FlamingPhoenixFighters, Marines, CafeMasters, BlindStars, JokerSlam.

At the same time, high-rated players who can (and most likely would like to play in Fischer) are here around, the only thing is that there is no a dedicated feature to help them coordinate and form a team.

We see that MlodyBog, marjohn, jfrl, mccannj, danijelo (covering the range of 1950 - 2200) play in Kasparov now, but I guess most of them would like to play in Fischer as well (and played there before).

So here below is a draft of how we can sort it out.

My proposal is to code a feature on TL website when a player who is going to be a captain, start forming a team by:
- defining the range of players' ratings applicable for his team, (for example, 1950 - 2300)
- and, perhaps(?), the range the final avg. team rating to be fallen within (for example, 2000 - 2200).

Not sure about the latter criterion, though, as the final avg. rating can not be defined unless four highest-rated players are enlisted, and once this formation is known while doesn't meet avg. rating requirement, it is not fair to kick out a player who was already enlisted.

Moreover, two criteria instead of just one make this query so strict that probability of team creation drops dramatically.

Ok, let's proceed with description of the enlistment process. Any player who meets this rating requirement and wants to be on the team, puts himself in the list. The list of players on the team is growing (and it is published on a dedicated webpage), and once 6 players are enlisted, the team is automatically created (or, another option, the captain can terminate enlistment process on his own, if for example 5 players is enough number for him).

I think coding of such feature is not complicated, it resembles the regular team submission process, the only difference is that the list of the team under creation should be public.

Looks like this feature could be used for creation of any team of any rating (the only thing is that the final avg. rating of the team falls within a wide range and can hardly be forecasted, so it's a low accuracy technique rating-wise). But this feature could help to bring several teams in TL action in addition to the teams created under regular procedure.

Would be interesting to know opinion of other captains on this subject.

Historically speaking, we really aren't seeing a narrowing of the top section. The last few years when we often had 6 or more very strong teams are the exception in the history of TL (or at least the 15 years I've been involved). Having 4 such teams was a pretty good result back in the fixed section days, and even for much of the flexible section era. When you consider that we had more than twice as many teams in those days, it's clear that we are doing comparatively well at the top end.

I think our main problem is just the decline of players overall. Comparing to historical circumstances, I see this decline being most evident not at the top, but at the bottom, and to some extent in the middle. I haven't checked on the lichess league since the beginning of the year, but they seemed to have many more players between 1400 and 1900, and not so many stronger players as we do. Although that might just be because rating systems aren't comparable between different servers.

Thanks, sbc, for the historical outline.
I have to mention that I am not a programmer, so my proposal is just an idea to be considered and developed by others, once (and if) it is basically accepted by captains and adopted TL admins.

Hi,

1. Just to clarify: earlier I described two potential systems - one used by lichess and another one completely made up by me during a single episode of TV series (so: it may make no sense). The lichess one doesn't assume random assignment of players to teams. The major difference between lichess and FICS is that they put all 20-25 teams into a single section of a Swiss tournaments. Therefore it is hard to be so outnumbered as is Mysterious Queen Moves this season. Plus there are the various smaller differences noted by pchesso.

2. I love the idea of pchesso that all teams should be rated average below 2050 (or below something). That is so simple. Of course, there is the disadvantage that one player couldn't play in more than one team in a single season (but maybe that change would make more players happy than unhappy...).

3. Now, I start to like the idea of fixed rating caps as I could just tell my captain "do not put me in a Fischer-section unless you can guarantee the team to be over 2100".

Michał

If you assume that most teams have a top board of around 2000ish then it means that in a section with a max rating of 1848.75 the lowest rated player they can take to the section limit is a player of around 1700 assuming that is their board 4. That means that it is very difficult to get any player under 1700 in lowest section unless you get 2 or more players to act as a counterweight to get team to max rating or you're willing to have a team substantially below max rating.

As a result it's not a surprise to me that when checking interested players list that Bardu (1369), Areckx (1291) and Yahnozha (1243) who are lowest 3 rated all didn't find a team as with current numbers it's very hard to put them into any viable team in contention for title. In fact the only way to get near the limit with just 1 of those players is to have three players rated 2000ish.

Problem is that as the bottom is getting cut off eventually the middle becomes the bottom as the section limits rise and suffers the same fate. I'll also note that in an effort to avoid the Fischer Section more and more teams are dropping stronger players into the bottom section that magnifies this effect. For example we have an average rating of 1996 on top boards of bottom section that increases to 2034 when you remove the weakest top board. Even scarier is that if you take top boards in bottom section and assumed it was a team it would be rated 2119 and be 3rd strongest team in Fischer Section.

P.S. For next season I'm planning to put 1 Fischer Team and then get all the other teams relatively closely matched. It might be something like 3 teams rated 1850ish or maybe something like one at 1900, one at 1850 and one at 1800. It means sections get narrowed but it's only way to avoid problems fixed sections prevent

As a result it's not a surprise to me that when checking interested players list that Bardu (1369), Areckx (1291) and Yahnozha (1243) who are lowest 3 rated all didn't find a team as with current numbers it's very hard to put them into any viable team in contention for title.

This is the kind of thing that makes me question whether we should continue with flexible rating sections. The system has made it harder for lower rated players to be placed in teams.

I have to mention that I am not a programmer, so my proposal is just an idea to be considered and developed by others, once (and if) it is basically accepted by captains and adopted TL admins.

Please keep doing so! This is exactly what we need the discussion board for. The structure of TL is not democratic, so we need the input of members here to allow new ideas to flourish; that's how we got flexible sections in the first place.

Anyway we have a good stretch of time to consider these matters, as I don't anticipate opening T77 until February.

I was hoping we could get back to the four-seasons-per-year scheme, as the larger breaks haven't translated into more teams registering, but I understand the staff appreciates their rest, especially after a full season (7 rounds + playoffs). With three seasons per year, I suppose tourneys should kick off in the first half of February, June and October, as in 2019?

Yes, our current intention is to work to a similar calendar as last year. Time off for TL staff is certainly a major consideration. On the plus side, this should give me more opportunities to update the history and champions pages, which Beto has been very politely reminding me about for some time!