webdiplomacy.net

FvAS Set 3 Damon Huntington - thuan0812: 2-0.

Phew, finally caught up on all of this (I think) :D All current sets should now either be in setup or active.

FvsAS Set 33 xorxes bt Unstupid 2-1

FvAS Set 2 Game 3
mhsmith0 2
Javatiger 1

GvIS Set 1 sereginss vs sw3dg!n 1-1

Going into a third game I guess. We both won as Germany

Can we think about something to not decide the winner over chance. Not for this one, but later ones maybe. The one who's picking Germany has clearly more advantage. Instead looking at something else. Sereginss suggested looking at supply center difference, might work but I think it can lead to dragging games to get more sc. Another way would be looking at what year you've won. Being quicker means you have a better offense and/or you have a better defense on the loosing side. Games finished on the same year, you can look at the sc difference I guess.

In this instant, sereginss won in 07 and I did in 09. Sereginss would continu, I'd go to the loser bracket. Seems fair, sereginss was the better player I feel

What are the stats for win rates for G/I? My thought was that it's substantially more balanced here than, say, F/A? Admittedly that could be off I guess?

I'm not only talking about G/I also for F/A. A quick look to past 1v1 games shows that in the sets with a 3rd game, the one who picks Germany, resp Austria wins the set.

Tiebreaker being who won earlier would be silly imo. It would discourage risky play which can make the games fun and different.

I woudl expect that the tiebreaker goes to whoever randed Austria at least like 80% of the time. I'm much less sure about it in G/I, I think that you're likely to see more balanced results though I could be mistaken? I don't go nearly as far as our TD and say that Italy should be outright favored, but there are definitely things you can do as Italy to give you a decent shot at the game.

Tom Bombadil wrote: ↑

Thu Feb 01, 2018 5:46 pm

Tiebreaker being who won earlier would be silly imo. It would discourage risky play which can make the games fun and different.

Can be true. I'm justing opening some discussing on the tiebreaker and how we deal with it.

If the tiebreaker goes 80% to a winner based on chance. Is it worth it to use this tiebreaker?

sw3dg!n wrote: ↑

Thu Feb 01, 2018 7:52 pm

If the tiebreaker goes 80% to a winner based on chance. Is it worth it to use this tiebreaker?

I think in the previous FvA Showdown tiebreakers went something like 51% to Austria, 41% to France, and 8% draws awarded to France, so pretty even. I suspect GvI could be a bit more unbalanced, but we need more data.

No side has an 80% edge in gvi.

So I've skimmed finished games and ongoing games in this tournament, and it looks like at no point has there been a split where Italy won both games. I think this lends some credence to the notion that Germany has a substantial edge; however, I think there's a pretty realistic chance that a lot of the edge is that Germany is simply an easier power to play, and that as people get more used to the nuances of the setup, you'll see something closer to balance (and if that's the case, you'll see a big drop in the German edge as you move forward in the tournament).

I still believe that Germany has a permanent edge on Italy, but that's just me. Italy is by far the lesser power among weaker players (which I'll lump myself into. I don't think I've ever won as Italy)

The 80% I took out the post above mine. I hope that's clear.

We'll see how this tournament continues. This is my first participation in a 1v1 tournament. So maybe I'm not seeing everything yet. It just stood out that in this tournament, the sets that ended in 1-1, the ultimate winner is the one who picks Austria/Germany. And that doesn't seem really fair to me

So looking at tourney spreadsheet from last go around (A/F)

https://docs.google.com/spreadsheets/d/ ... edit#gid=0

Round 1 winners bracket France won an outright majority of the tiebreaker gmaes, and got an additional +1 for a game that was drawn (since draws go to france)

Round 2 winners bracket Austria went 6-3-1 so effectively 6-4

Round 3 winners bracket Austria went 3-3

Round 4 winners bracket two draws, so France effectively went 2-0

Round 5 Austria 1-0

Round 6 Austria 1-0

And the semis and finals were outright sweeps by xorxes

But overall this seems... about balanced for results?

I think we'll see how things go for I/G, but it strikes me as the sort of thing that is either unlikely to be a large problem, or likely to be solved in next year's tourney. FWIW I'd guess there are a couple of people (xorxes most obviously, probably a couple others too) who are basically going to sweep their way through both tourneys (and yes, xorxes went 2-1 instead of 2-0 in A/F but still), so my guess is that the tiebreak impact will be relatively minor overall. Obv could be wrong but that's my current guess.

G/I seem to be imbalanced according to vDip stats, though not wildly so. Out of almost 1500 games played, you're looking at 60% Germany, 5% draw, 35% Italy. It's still enough to make me worried, at least.

http://vdiplomacy.com/stats.php?variantID=23

Vdip stats are low level. Ck. F/A: 2k games, A-55%, F-39%, D-6%...

Just FYI everyone, have had a busy few days and that's why some of you are still waiting on setup/join PMs and tournament bracket updates. I'm planning to get through them all tomorrow for both tournaments.

If you've sent me a message and I haven't responded yet, please don't send me another unless you sent it to me before Wednesday. My inbox is struggling already and getting a lot of duplicate PMs can flood other ones off the page.



With regard to the tiebreaker method - I'm definitely not changing it during the tournament. If there's a significant problem with it, I will definitely consider alternatives for next year's Showdown(s).

The reason I chose the tiebreak method I did (the extra game) is that these tournaments are a test of your ability at playing the 1v1 variants as they are on the site - as such, the strategies you use should be close to being identical to ones you would use if you were playing a regular, non-tournament 1v1 on the site (against an opponent of the same level).

Unfortunately, some compromises have to be made to fit the tournament format - for example, you can play for a draw as one power provided you're certain you'll win your other game - but wherever possible, I've tried to avoid incentivising strategies in this way. Due to this, making the tiebreaker something like SC Count or Year of Victory was never an option - these massively incentivise certain strategies, with an SC count tiebreaker encouraging players to aggressively strike and cripple opponents so that they have the freedom to position around a large number of open SCs prior to ending the game, and Year of Victory massively encourages rushes (and also favours players who are more comfortable on the nation with the better ability to rush). The only tiebreaker that kept in with the idea of the tournament that I could see was a third game, and although again there had to be a compromise - allowing one player to play for a draw there - it was absolutely necessary for set duration and it only affects the third game, not the first two.

I hope that all makes sense - fairly tired rn so it may be a bit jumbled :D

If GvI proves to be massively imbalanced, I will likely return to just running the FvA Showdown next year. However, I don't think we've seen enough competitive play to say that it's imbalanced yet.