League General Discussion Thread

gamerman01

And I see you have found just how convoluted they are and why they should not be used, if I dare say so.

gamerman01

To be fair to the BM guys, who did a great job,

Larry’s flying tigers are also a kind of lame attempt to make things a bit more historical. The flying tigers never supported Chinese infantry, of course. They basically did strat bombing runs from what I recall researching it a long time ago

gamerman01

And then that special unit/situation necessitated further rules, such as the fighter can’t go out over water.

gamerman01

Of course the intent of Larry Harris, in the case of the flying tigers and kamikazes, and also the BM team was to make the game MORE FUN! and they pretty much succeeded. But those Vichy rules…

mainah

Thank you for the explanations and the time. Really appreciate it.

And yes, way clunkier than the Mongolian rules.

They did do a great job, it’s not easy writing rules. If it was, there would not be a profession called lawyer.

I am not saying that they shouldn’t be played. Just clarified.

gamerman01

You are close to the point where you can help other league players understand and enjoy them.

gamerman01

I just got ambushed by it in the first round of the playoffs. I hadn’t played many games for years and got blindsided. Lost 2-3 UK air because I landed them on Syria and not an island. Just brutal.

mainah

@gamerman01
I lost 2 fig and a tac in french indo-china, same thing.

gamerman01

Alright, now I can’t resist complaining about those rules.
Germany goes out of their way to take Normandy G1. Then the UK can stop it with a transport to South France or Normandy. Depending on circumstances, Italy can activate the whole thing with that tank they have in North Italy. In this case, Italy’s combat movement has to be made not knowing the outcome of that battle (There is a French fleet out there) It’s so messed up. In my opinion.

gamerman01

@mainah said in League General Discussion Thread:

@gamerman01
I lost 2 fig and a tac in french indo-china, same thing.

Ah, because you were sure there would not be a J2 DOW I guess.
Interesting

mainah

@gamerman01
Had an opportunity to UK1 DOW on Japan to sink a japanese fleet (1bb 1 CA 1dd 2 TT) and took it. Probably not for the best.
Between that and an unforced error costing me a few uk tt’s I think avner has this game in the bag, and it’s only at the end of round 3.

gamerman01

Oh, right, had not considered a UK1 DOW to hammer a vulnerable Japanese fleet!

Stucifer

Wanted to double check a rule here for BM/PtV:

In OOB you can attack USSR with Italy and NCM German units in on their turn giving you the NO for Trade still, but being in original USSR territory. Is that legal for League play?

MrRoboto

Yes it is

gamerman01

Boom a thought just hit me.

If anyone is wondering about “grade inflation” where we have a lot more E’s than 2’s and 3’s this year, I think I’ve suddenly realized a major factor.

Rather than starting everyone out at 0 each year, the ranking system carries over everyone’s reputation from the prior year, everyone who finished 3+ games the prior year.

So we start the year with a lot of 1’s, E’s, M’s, and those players are much less likely to leave the league because they’re crushing it. The weaker players lose heart and go back to beating their neighbor kid next door!

It’s all relative anyway, so it’s still good. (e.g. a 4.50 is definitely significantly higher than a 4.00) But it did bother me that we haven’t had a central distribution surrounding 4.00, middle of tier 1, and I’m pretty sure (without even thinking for 10 minutes) that that is why.

I post it here so you guys can kick it around a bit and give thoughts.

MrRoboto

This probably explains the inflation of points.
However, this is the least of the flaws the current system has. While I do appreciate that the current system is a vast improvement over what was in place before (a simple Win%, nothing else), I still have numerous issues with it.

No definite points after a result.
Imagine the following scenario:
Player A (Tier 1) wins against Player B (Tier 1).
You could score Player B first. He lost against a Tier 1, so receives 2 points, is at 2.0 PPG and is therefore dropped to Tier 3. Player A now receives 4 points for winning against Tier 3, is at 4.0 PPG and stays Tier 1.
Or you could score Player A first. He won against a Tier 1, so receives 6 points, is at 6.0 PPG and climbs to Tier M. Player B now lost against a Tier M so receives 4 points, is at 4.0 PPG and stays at Tier 1.
If you score winners first, you have points inflation.
End of year standing actually reflects yearly average
Your end of year PPG is actually a yearly average and not a reflection of your skill at the end of the year. Have you improved over the course of the year? Hard to tell, your skill might be Tier M at the end, but you still score at Tier 1.
Not predictable
If I play against a certain player, I don’t know how much that game is worth until the 31st of December. My opponent can climb or fall Tier until the end of year and will retroactively change my PPG. That will even happen without my interaction with said player! I might stop play in June and my PPG might severely change until the end of year, which is very unintuitive. And I don’t think it makes a lot of sense.
Discourages playing weaker opponents
Everyone who has played Tier 3 or Tier 2 has lowered the PPG, even with a win! We are talking about 50+ games this year alone.
As long as your PPG is higher than 4.0 (This is only Tier1, the MIDDLE of the pack, not elite players!), EVERY win against Tier 3 players is hurting you!
As long as your PPG is higher than 5.0 (Still not the highest Tier, only Tier E at this point), even winning against Tier 2 is detrimental.
And when your PPG is higher than 6.0 (All Tier M, but even some Tier E players!), you shouldn’t even play against the middle Tier, Tier 1. Even a win would hurt your PPG.
Losing can help
Every Tier 3, Every Tier 2 and even half of Tier 1 can easily improve my just losing to a Tier M.
Sorry, but this makes no sense
Results against new players depend on a single moderator
Currently, new players are not ranked consistently. As an example: @jkeller r is 0-1 but placed at Tier M. So @AndrewAAGAmer received 8 points for that win.
On the other hand, @Gorshak is 2-0 against two top players, @666 and @GeneralDisarray but still placed at Tier 1. So their losses against him gave both of them only 2 points, DRASTICALLY lowering their PPG.
Circular referencing
A players Tier affects the points opponents get. These points affect the opponents PPG and thus the tier of the opponents. This Tier affects the points the original player gets and thus the Tier. But this Tier originally affected the points the opponents get.
This goes back to issue 1, but the problem is more widespread and generalized than the example in 1).
You can reach dozens of different rankings of every single player here, with the exact same game results, depending on the order of calculations and the order of reporting.

There are a few minor things that theoretically could also get fixed, but those 7 points are major flaws with a PPG based system.

I have created an improvement, which is ELO based (a system used in games like Chess, World of Warcraft or League of Legends). That fixes all of the issues above and the spreadsheet is already finished too, fully automatic even.
I’m just waiting for feedback from @gamerman01 before I share it with all of you

Ghostglider

Moving towards an ELO based system sounds like an improvement to me.

Martin

@MrRoboto thank you for this analysis, and well presented! The ELO system was suggested a few times during the last years, and I also strongly support it. And as I stated before, there are plenty of management systems available which would ease the job of the score keeper / league manager.

AndrewAAGamer

I played at Days of Infamy and they used an ELO system. While I do not understand the finer points as you obviously do the best thing about it was you could play anyone; the points earned or lost reflected that. As a top player I would only gain a minimum of points, and they would lose a minimum of points, for beating a bottom player and yet if they could score a win they gained a ton of points and I would lose a ton of points. Therefore, there was no discouragement to playing anyone as there is here.

Plus, it sounds like it would greatly alleviate the work load for the League Moderator which is a very good thing.

jkeller

I don’t really appreciate being used as a counter example here. I decided to take a break this year, but if you look at my results from last year you would see that I clearly belong in the M tier or you can ask any of my opponents. AAGamer absolutely deserves every one of those points.

As to the entire argument I have no issue with an elo rating system in general, as I am a lifelong avid chess player (Master there too fwiw) and it is generally pretty accurate. I would say that due to the length of time it takes to finish one of these games, and the resulting low sample size the ELO rating would be far less accurate because it does rely on a larger number of games to be accurate. In other words AA players just have too few games elo rating to be accurate.

League General Discussion Thread

Featured Topics

T-shirts, Hats, and More

Suggested Topics

L25 OOB elche (X) vs. Martin (L+35)

PTV - regularkid (Axis +8) vs mikawagunichi (Allies)

L25 PTV Stucifer (X+11) vs Axis-Dominion (L)

L25 OOB TSK23 (X) vs Martin (L+40)

L25 BM4 fasthard(X) vs BombsAway(L+21)

crocket36 (Allies) vs Simon33 (Axis+6) French Rebalance

L24 BM4 Amon-Sul (Axis) vs Fasthard (Allies+24) - Game 5

L24 PtV Gorshak (Axis+10) vs ArtofWar1947 (Allies) #2

39

17.7k

40.4k

1.8m