Are 1vs1 statiscics collected shomewhere?

[Quetzalcoatl]

the idea (randomize maps plus position) is there already, SOMEONE just has to enforce it - u got my vote to do that, and i dont care starting from scrap. would make the competitive system more credible.

Ah well is it so important? Competetive playing is much less fun if ppl cant win So... Having rules that make winning harder would make it less fun

[Doc Paterson]

the idea (randomize maps plus position) is there already, SOMEONE just has to enforce it - u got my vote to do that, and i dont care starting from scrap. would make the competitive system more credible.

That would be awesome, and I think the ladder would become a much more accurate ranking of skill if it were implemented.

Isn't the ladder severely lacking in administrative support? I vote for Rigor and Q seizing power and initiating their bold new system.

At the very least, such an attempt would cause others to step up and take a more active role in things. Right now, the ladder has what, one admin? (TMOB) There's a good chance he'd be in favor of this system, too.

Seriously, rush the mostly-empty capital building. I'll provide cover-fire.

[Quetzalcoatl]

I remember that idea was discussed in the past in ladder topic. And grrr made this cool random addon after that. However too me it looks that ppl are not realy interested (or rather dont care at all) in changing the way things are right now

[Yogibear]

I remember that idea was discussed in the past in ladder topic. And grrr made this cool random addon after that. However too me it looks that ppl are not realy interested (or rather dont care at all) in changing the way things are right now

*Yogi is interested and cares, he just doesn't have much time to play ladder atm*

[Rigor]

thanks for your support, i was asked already to do that. the answer is: with pleasure!

[grrr]

*Yogi is interested and cares, he just doesn't have much time to play ladder atm*

And it should also be stated that Yogi fixed some bugs regarding MP campaigns, but only for 1.8. IIRC the timer stuff was fixed (haven't tested it, sorry), which is the biggest drawback for the RDM add-on on the 1.6.5 server.

[Quetzalcoatl]

I thought it may be worth updating results I posted before as 1.8 came out few days ago. So here are the results from replay archive start till release of 1.8 (10037 games found). Looks like there are again some issues with timer synchronization between ladder and archive websites in last few days (28/29 march+) but I'm not going to fix it only to have 20-40 more games inspected.

I will try to transform this crappy script to something more usable that everybody will be able to use easily, but don't know when I will be able to do so as I'm insanely busy lately. If someone is interested in participating in this initiative please drop me msg.

[Weeksy]

This data indicates two things that I find a little concerning: The player 1 advantage seems significant in quite a few matchups, and, at least at the 1800 level of games, northerners v. knalgans seems biased towards the northies side. Now, I know that this isn't a lot of data, and there might be various reasons why it is skewed this way, but still, these results are not optimal.

[Rigor]

thats right, we got tons of data supporting that theres just something wrong with northerners, but who cares.

all games ever played on the server when the replays were archived

----------------win%---------------------win%---------------
Drakes (P1) 161 61.92 Northerners (P2) 99 38.08
Drakes (P2) 114 39.45 Northerners (P1) 175 60.55

-------------------------win%----------------------win%---------------
Knalgan Alliance (P1) 133 50.00 Northerners (P2) 133 50.00
Knalgan Alliance (P2) 92 31.83 Northerners (P1) 197 68.17

------------------win%---------------------win%---------------
Loyalists (P1) 205 59.08 Northerners (P2) 142 40.92
Loyalists (P2) 127 38.02 Northerners (P1) 207 61.98

---------------------win%----------------win%---------------
Northerners (P1) 257 65.23 Rebels (P2) 137 34.77
Northerners (P2) 134 47.35 Rebels (P1) 149 52.65

---------------------win%-----------------win%---------------
Northerners (P1) 189 62.17 Undead (P2) 115 37.83
Northerners (P2) 141 45.19 Undead (P1) 171 54.81

---------------------win%--------------------win%---------------
Northerners (P1) 135 59.47 Northerners (P2) 92 40.53

early game all players

---------------win%--------------------win%---------------
Drakes (P1) 75 60.48 Northerners (P2) 49 39.52
Drakes (P2) 50 31.45 Northerners (P1) 109 68.55

------------------------win%--------------------win%---------------
Knalgan Alliance (P1) 56 49.56 Northerners (P2) 57 50.44
Knalgan Alliance (P2) 33 22.92 Northerners (P1) 111 77.08 <------------ wth

-----------------win%--------------------win%---------------
Loyalists (P1) 86 60.14 Northerners (P2) 57 39.86
Loyalists (P2) 44 28.03 Northerners (P1) 113 71.97 <------------ wth

--------------------win%--------------win%---------------
Northerners (P1) 11 84.62 Undead (P2) 2 15.38 <------------ wth
Northerners (P2) 4 57.14 Undead (P1) 3 42.86

--------------------win%-------------win%---------------
Northerners (P1) 17 80.95 Rebels (P2) 4 19.05 <------------ wth
Northerners (P2) 6 66.67 Rebels (P1) 3 33.33

-------------------win%------------------win%---------------
Northerners (P1) 0 0.00 Northerners (P2) 9 100.00 <------------ wth

what i think deserves attention as well is the fact that drakes seem to be a good race to fight knalgans, did u notice that?

early game

--------------win%-----------------------win%---------------
Drakes (P1) 5 62.50 Knalgan Alliance (P2) 3 37.50
Drakes (P2) 6 66.67 Knalgan Alliance (P1) 3 33.33

late game

--------------win%-----------------------win%---------------
Drakes (P1) 4 80.00 Knalgan Alliance (P2) 1 20.00 <------------ wth
Drakes (P2) 4 80.00 Knalgan Alliance (P1) 1 20.00 <------------ wth

looks like after u survived for 20 turns with drakes it doesnt even matter if u r player 1 or 2
there is also data where u have drakes with a lower victory count but taking all games into account this seems to be rather seldom the case:

all games


----------------win%-------------------------win%---------------
Drakes (P1) 170 69.11 Knalgan Alliance (P2) 76 30.89 <------------ wth
Drakes (P2) 99 42.13 Knalgan Alliance (P1) 136 57.87
Drakes (P2) 99 42.13 Knalgan Alliance (P1) 136 57.87

[grrr]

Can we try to get the same data for +1800 and 1.6 revX maps? I'd like to compare the FMA on the 1.6 maps vs. the FMA on the maps from the map updates add-on. It should tell us how much Doc's village hack helps.

[Caphriel]

It may be worth noting that a disproportionate number of the games recorded, especially amongst the higher ranked players, are played on Freelands.
Edit: Or at least this was true for the first set of stats, and I don't see any reason for it to have changed.

[Doc Paterson]

This data indicates two things that I find a little concerning: The player 1 advantage seems significant in quite a few matchups

We've already talked about this. You should read the earlier sections of this thread. I'm not saying that there aren't varying degrees of P1 advantage on some maps, but we need to remember when viewing this data that a vast majority of the time, player 1 is not only choosing the map, but is choosing a map that they're familiar and comfortable with. That is hugely significant.

I'd be extremely interested in seeing data from grrr's random 1v1 option, though I'm not sure it'd be possible to single those out....

[Rigor]

yes thats right. a game where a rank is established based upon the intention to play at this specific time, be p1, and pick map and maybe even fraction certainly has a HUGE impact on your score even though your maybe a crappy player overall.

i think in a competitive system this should not have space at all especially not because we all know that a rather noobish player can be intimidated by those high score players and might contribute in fortifying their position even more!

one thing you should consider when looking at % is always the sample taken, for instance only 10 games shouldnt be representative for the overall situation, as well as considering that 50% of the time the map is freelands in those samples representing high score players (30% overall for all games played). im saying that because of my own example taken but unfortunately there is not more data - in 10 000 games actually a joke one might think

[Quetzalcoatl]

About northies faction balance its probably also worth to mention that they are considered to be quite easy faction to play, so many ppl who like to pick map/faction often chose them.

Can we try to get the same data for +1800 and 1.6 revX maps? I'd like to compare the FMA on the 1.6 maps vs. the FMA on the maps from the map updates add-on. It should tell us how much Doc's village hack helps.

Done, however its not much games in total and most of the matchups doesn't have enough played games I wasn't expecting that there will be so few results for updated maps

I'd be extremely interested in seeing data from grrr's random 1v1 option, though I'm not sure it'd be possible to single those out....

That would be kinda cool, but I also don't know if its possible to check that replay was played using RDM. There are some problems anyway even if that would be possible. First of all I downloaded replays from archive and RDM games dosn't save have replays in archive (don't know why this happening ). Besides that even if there would be replays we probably would get small sample (like for updated maps) that wouldn't tell much as grrr released RDM not long ago and not all ladder players were using it

[grrr]

...

I was honestly expecting more, too ... anyway - thanks for your fast reply!

I'd be extremely interested in seeing data from grrr's random 1v1 option, though I'm not sure it'd be possible to single those out....

That would be kinda cool, but I also don't know if its possible to check that replay was played using RDM.

one could scan for

Code: Select all

scenario="rdm*"

inside the replays. It's not reliable of course, but then again I haven't seen that many people who label their games "rdmSomething", either =p
Factions would need to be guessed from the recruit list, but that's not too bad for default factions, at least. There were some more issues that were already spotted by Sorrow, I am not sure what the status of his project is (game-parsing-related. too).
I wonder whether one could simply use the campaign stats tool for the RDM replays, hm ...

Besides that even if there would be replays we probably would get small sample (like for updated maps) that wouldn't tell much as grrr released RDM not long ago and not all ladder players were using it

(OT: Well, we - as a group of ladder supporters, even if for very different reasons - have the power to change that, at least for 1.8 ... but one of the active players here has to step up as ladder admin first. See the ladder thread.)