2015 League General Discussion Thread

So let’s say someone attacks an sz with multiple subs.

Can you submerge 1 sub and retreat the other to another sz?  Or does the decision to retreat precede the decision to submerge, meaning that both subs have to either retreat or submerge?

Here’s the game thread for reference.
http://www.axisandallies.org/forums/index.php?topic=36650.0

With regard to the ratings……one thing I learned via playing on the TripleA ladder is that the most accurate A&A ratings are obtained by making each match-up a 2-game series where both players play both sides at the same bid.  In the 1941 anniversary ladder, it was necessary to be an expert at both sides to top the ratings.

It seems to me that Gamer’s ranking system is about as accurate as we can do short of adopting the model of 2 games with different sides at the same bid.

Actually what Zhukov just said sparked something I’ve been thinking about for a while about how we do bidding and what it means.

My understanding is that the primary purpose of bidding is to balance the game, i.e. offset the inherent advantage the axis has in the base setup. But the way we do bidding actually introduces a second purpose: handicapping the players based on how good they are with a certain side. We’ve all seen comments to the effect of “I would never play X at a bid lower than Y”, which is a sign that this also happens. No matter what, different players have different ideas of what a reasonable bid is, which is influenced both by their own relative strengths with the sides and their opponent’s.

Personally I’ve been spiralling farther and farther from ever playing with the Allies - as I play more Axis games and get better at playing the Axis, the bid I’m willing to accept increases, so I keep playing more Axis games, etc.

Now, I’m not necessarily saying that this is a problem or that it should be changed, I just think it’s interesting from a game theory perspective :)

It might be fun though to have a league standard bid (+25 or +30 or whatever people think is reasonable) and have the sides decided by a dice roll at the start of the game. I like the 2-game match idea as well, but it might not be practical with the length of time it takes to play global games.

And I’ll just add one comment regarding the rankings. It’s important to remember that in our rating system, and indeed in all rating systems, the actual rating number (PPG, elo rating, whatever) is virtually meaningless. The important thing is the ranking, i.e. where you place in relation to everyone else. In our system, the tiers are completely arbitrary, so it is (statistically at least) meaningless whether you are the lowest E or the highest 1.

@Zhukov44:

So let’s say someone attacks an sz with multiple subs.

Can you submerge 1 sub and retreat the other to another sz?�

Yes, this is possible.

Or does the decision to retreat precede the decision to submerge, meaning that both subs have to either retreat or submerge?

Here’s the game thread for reference.
http://www.axisandallies.org/forums/index.php?topic=36650.0

This is covered on page 19 of the Europe rulebook.

Submerge is done on step 2.  Press attack or retreat is step 6.

So every round, including round 1, the attacker and defender decide whether to submerge submarines before dice are rolled, of course only if there is no enemy destroyer present.

Dice are rolled

Attacker decides whether to retreat.

So the attacker could submerge any number of subs each round, at the beginning of the combat round.  At the end of the combat round, the attacker decides whether to retreat.  You can only retreat once, to a single sea zone you came from, but it is possible to leave any number of subs in the combat zone.

Good ideas, guys, as always

I would encourage you to implement your ideas with the people you play.  Propose a 2 game set where you take each side if you want.  Propose that you agree to a +25 or +30 bid and that you roll for sides, and play one game.  Remember that you can do just about whatever you want if you and your opponent agree.

Nielsen, I would suggest that if you want to play the Allies, you take them against a somewhat weaker opponent until you get more comfortable and honed like you are with the Axis.  You don’t need to stop playing the Allies, just don’t take them right now in a game you really want to win, with someone who might be as good as you.

I do think that the customized bidding from game to game between players is ideal for the most part.  Again, you can do whatever you want if you can get someone to agree.

I would say the primary purpose of the bid is this: it gets both players to agree to play each other, and makes each of them satisfied with the side that they have, given the bid.  It’s not necessarily to balance the game perfectly.  Maybe I just want to play the Allies against you this time, so I will keep bidding them down until you give them to me.  At some point you’re willing to give them to me, and then we’re both happy.  THAT is the primary purpose of bidding.

To follow up on my last thoughts about the rankings, keep in mind that the order in which the results are recorded actually matters for the PPG and tiers. This makes it difficult to determine what PPG you should have solely on the basis of W/L against various tiers.

As a simple example, consider the first recorded result of a season.

Situation 1: Both players start as tier E.

Winner gets awarded 7 points, loser 3 points => Loser gets bumped down to tier 2 => Winner’s points get bumped down to 5.

Result: Winner is tier E with 5 points, loser is tier 2 with 3 points.

Situation 2: Both players start as tier 2, winner’s tier gets adjusted first.

Winner gets 5 points, loser 1 point => Winner gets bumped up to tier E => Loser’s points get bumped up to 3.

Result: Winner is tier E with 5 points, loser is tier 2 with 3 points.

Situation 3: Both players start as tier 2, loser’s tier gets adjusted first.

Winner gets 5 points, loser 1 point => Loser gets bumped down to tier 4 => Winner’s points get bumped down to 3.

Result: Winner is tier 2 with 3 points, loser is tier 4 with 1 point.

Situation 4: Both players start as tier 4.

Winner gets awarded 3 points, loser 0 points => Winner gets bumped up to tier 2 => Loser’s points get bumped up to 1.

Result: Winner is tier 2 with 3 points, loser is tier 4 with 1 point.

So just with these 4 situations we’ve found two “stable states”, and two different ways of getting to each of them. Situations 2 and 3 even have the same starting point, but get different results depending on what order the tier adjustments are performed in. I haven’t included the last stable state for this setup (winner at tier 1, loser at tier 3), since it’s just symmetrical with the given situations.

Now consider what happens with the exponential increase in complexity caused by having hundreds of games. There are probably all kinds of various “stable states” we could find in the current rankings by shifting blocks of people up and down between tiers.

All of this is just to underline my earlier point: the PPG and tiers are arbitrary and don’t matter. What matters is your ranking compared to everyone else, which will not change significantly regardless of how the PPG and tiers sort out.

I’ve thought about this extensively when consulting with someone who wanted to use this ranking system in their own league.

Are you forgetting that I assign players to a default tier 3 for a few games?  Tiers are not decided from game #1, which is when it would matter which person you calculated first.

I could be wrong, I’m not 100% certain, I guess my pea brain can’t quite wrap all around this, but did you take into account that your tier is set from game #1, game #1 does not decide your tier.

Yeah, my examples assume that tiers for returning players carry over from previous seasons. I was trying to present the simplest example possible to demonstrate the concept, with the understanding that it also applies when extended to many players with many games.

And again, to reiterate, I just see it as a quirk and not a flaw in the system, as the ranking itself is always accurate. We just shouldn’t put too much stock in the PPG and tiers, as they can change for unexpected reasons. In other words, it’s your PPG relative to other players that means something, not the number itself.

ah well, at least i can claim i’m higher ranked than gargy  :-P

Well thanks for helping to calm axis-dominion down, bmnielsen

“Hmm 91.1% Axis win ratio in tier E…”

Axis win in tier E vs. tier E games is 75% (6 out of 8 so not that many games available for statistics) - Maybe bids needs to go even higher…

I agree.  Bidding’s not enough.  Somehow it always gets brought down to the 20-30 range we’ve all sort of settled on.

The Allies need some sort of “burned in” advantage on top of the bid.

I am starting to wonder if it’s because most players want to “win” the bid by having the last bid.  What would happen if we bid to take the Axis - probably the handicap to the Allies would get bigger?

Remember, you guys can bid however you want - no one says you have to bid for the Allies, start high and go down.

I know some players are vehemently opposed to this for some reason, but another way is to take stuff away from the Axis.  Might make for more interesting starts - the Allied bids are getting to be so standard now some of you are probably bored with them.

Try taking the complex off Ukraine.  Or, since Larry used to have a minor in Berlin for some unknown reason, try keeping the major in Berlin but go with a minor in West Germany?!!

Anyway, I invite you to get more creative - if you and your opponent can agree……

I guess I’m saying several things, but one is -

Bidding a lot of extra units to the Allies is only one of many ways to make both players happy with the starting setup.

And yes, pretty much everyone is settling for not quite enough boost to take the Allies and the results show it, just about any way you look at them.

I’m truly not sure a bid ever balances this game out.  Almost all of the other Axis and Allies games can be balanced out, but this one just doesn’t look like it can be – is that what we are seeing?  I mean, there comes a point where a bid just simply overpowers an area of the board, and I think why the bids are in the 20-30 range is because any more than that and one of the game theaters becomes almost completely broken.

Just to throw out an example, in 3rd edition (CD play), 21 was probably the most even bid.  However, because someone could plant 7 units in Ukraine (“Power Europe”), it imbalanced the game to such an extent that two things happened.  One, bids over 18 were rarely given (sometimes 20, which was still just 6 land units) – therefore at the top level the Allies always had a slight superiority that could not be completely balanced out (but unlike with Global, that was like a 52/48 advantage, so very slight iirc).  But this type of bidding also allowed things like Power Africa or Power Asia – in essence, an entire game theater had to be changed to balance out the game.

Now, here is the difference with our league.  Because the Axis only have to win on one board, I’m not sure the bid can balance everything out – Axis can just go “where they ain’t” so to speak.  I don’t remember this problem with AA50 because the winning conditions were for the entire board.  And this is true of all of the other Axis and Allies games that I can think of.  So, instead of being able to get it close (such as a 52/48 spread), we will never be able to get it that close without having bidding parameters on both boards or other things that will just make the entire bidding process cumbersome.

Too bad we can play good ol’ AA50 in league anymore lol.

On another (and probably better) note, there is really an easy way of balancing out who bids last, and so forth.  You do it blind.  Did anyone here ever play 2nd edition PBEM at AAMC?  You would submit your bid to the bunker without knowing your opponent’s bid, and the low bid “wins” Allies (back then the bid was for Axis).  In the case of a tie, a virtual coin would be flipped for who gets what side.  Would it be that hard for someone to program this process?  I doubt it.  But this would only be if everyone is on board.  The true purpose is that you only get to bid once (so make it good!) per game, so no bidding down your opponent back and forth (which I find kind of annoying anyway).

To summarize, blind bidding via a “bunker” and having better bidding parameters are my suggestions for improvement.  Changing the entire layout of the board and other rules just seems to be too much for a “league”, since again the purpose of league is to play a standardized game (imho, which is not shared by all).

Fabulous, thank you so much

You are aware many players play with a limit of 1 unit per territory for the bid?  That helps a lot.  I think we can go toward 30 without wrecking the game, when this limitation is in effect.

To a certain extent, I believe we have exasperated the balance problem with our rules, which perhaps were simply adopted from previous years? IMO, all the rules under section 4B need to be eliminated. This will allow for diversity to the bid and perhaps help the extreme unbalanced nature which currently exists.

Agree - we will be re-examining everything in light of what we’ve learned in 2015

Alright, I just thought of something I need to make clear.

A lot of you guys, including and especially Axis-dominion, have demonstrated a misunderstanding of the purpose of the tiers.

They are not status symbols.

They are a tool I use to rank players considering the strength or weakness of your opposition.  It was never meant to be an exact science.  The tier system was developed to be superior to straight win-loss records.  Win-loss records alone strongly encourage everyone to play weaker players, but obviously everyone can’t play weaker players.

Naturally, over time that is forgotten and lost, and players will naturally start gauging their progress and perhaps set personal goals based on PPG and tiers, but that was never the purpose or intent for them.  The purpose of the tiers and the points is so that you don’t get the same credit for beating the best player in the league as you get for beating a newbie.

Bmnielsen is right, it’s about the ranking #, not the PPG or tier.  The PPG and tier are means to an end.  So all this hoopla about losing tier E status…… well, think about it in light of what I just said, please.  :-)  :-)