How consistent are you in TWD?

[Weaver]

So I was thinking it would be neat to have a stat that shows how consistent you are. Something to do with your deviation from your "average" game perhaps..? I'm not much for statistics, but maybe someone around here is. Anyways, just some food for thought. I'd be interested to hear what others have to say about it!

[Lofty]

Milosh, or whoever works on the TWD bots, here's two propositions -

# 1

1:LofTy> I have an interesting proposition for thinning out more accurate averages on TWD site
1:LofTy> Perhaps it should be made that anytime you win against a squad with a rating or ave rating higher than say, 1250 or so, it sets aside a seperate kill/death win/loss ratio column
1:LofTy> Than any times you play against a squad lower than that
1:LofTy> So you'd have two columns
1:LofTy> Might not be what you were thinking of integrating, but would still help see who's actually good against stronger squads and who just plays against weak ones always
1:LofTy> There are a lot of players on weaker squads that have very higher k/d ratios, but that's because their squad overall does not even play anything past low caliber squads, and so its hard for recruiting scouts to tell if those players would be worth anything against higher squads without extra digging in TWD records

# 2

1:LofTy> Well, an interesting thing that could be added as a counter column to support your idea
1:LofTy> Is a 'choke count' column
1:Weaver> hmm true
1:LofTy> It adds a point for every round you go out or get subbed -4 or higher
1:LofTy> or higher negative I mean
1:Weaver> yeah something like that may be interesting
1:LofTy> Then out of that, a choke average can be derived as to out of how many games, are there choke ones
1:Weaver> although.. you might want to see it as a % of games played
1:LofTy> Yeah, what I just said up above yours
1:Weaver> yeah
2:top> sorry matt
1:LofTy> That can be a % ave column next to the choke count
2:Mattey> all good holmes, i blew it too
2:Hercules> talking bout ur load
1:Weaver> yeah that wouldnt be too hard to do

One, or perhaps both ideas could be intergrated into the stats system on the site..I think it would be beneficial with breaking things down more, and seeing who are more, or less solid players.

[Money]

wont work

[Soup du Jour]

... Something to do with your deviation from your "average" game perhaps..?

perhaps a standard deviation? :P

If the can code so the bot can find averages, STDEV should be pretty easy too.

[Pressure Drop]

It wouldn’t really mean that much as a lot depends on the quality of the opposition and if you mean average game I’m going to have to think about your k:d ratio.

So if you constantly play weaker teams it would give a higher kill ratio with a low standard deviation

If you played any team that’s willing, no matter how good you are that is going to result in a higher standard deviation. Yet that isn’t a true reflection off inconstancy in the player more so the inconsistency of the opposition

Weighting it in terms of the league points the opposition have also would throw up problems because lots of squads have varying lines and some triple league squads will have people playing a league they don’t normally play just for fun or a body is needed etc … so that is handicapping some talented players and providing easy kills vs a decent rated squad for their opponents.

Still if you added an average rating column for the average league points of the squads played vs like there is in the league it would add a bit more objectivity to the K : D ratio even though the flaw above still applies.

You could try to weight it with a kind of individual ELO within the team league points

Like

(Player1Rating + Player2Rating + Player3Rating + Player4Rating)/4 – (opposition1Rating + opposition2Rating+ opposition3Rating +opposition4Rating)/4

Keeping track of the individual performance and trying to see which line should theoretically win and get a % win chance then weight your individual score vs your K : D performance in the match and adjust it for a decent performance.

So if you are in a terrible squad and you go 8:10 against a good with a good line in, it should increase your score depending what your score is already etc …

Maybe there is a simpler way to do this, as I see it would be complex and still be hard to get a balanced stat that actually meant a lot and would reflect your performance – specially with all the squad hopping and name changing

[Lofty]

Uh pressure drop..that's why the choke counter could be split two one for each of the two columns I spoke of in the first log up there...where it ranks your statistics seperately for squads rating 1300 or higher and all squads played under that, for instance. Hence...it would be a lot more accurate.

[Pressure Drop]

Well lofty what you said comes from an elitist point of view and solely for DD

In JD there are only 3 squads above 1300 and that’s towards the end of TWL qualification too, in DD there are 9 squads.

What you are saying is draw an imaginary line at 1300 and take the stats separately so newbs can have their crappy column and you can have a more accurate one for the TWL squads …

Well that’s what TWL is for, to see how the top lines perform against the top. Apply more meaningful “choke” or other stats there.

But by drawing the line at 1300 in TWD the people you affect most are those in the squads 1300 +/- 75 or so. So if a squad is borderline 1300 play a weaker line by their own standards vs the top TWL squads strong line and get well beat 50 -36 in a 4’s match each player turning in a 6:10 do they all really deserve a choke point added?

Not to mention the fact that it won’t apply to all of the TWD season so before the teams break 1300 ppl could acquire choke points in there <1300 column then playing a month later with the exact same performance vs the exact same line up, acquire one in the >1300 one. Or subbing at -2 or -3 and avoiding a point to skew the stats

You also haven’t addressed what exactly happens to a team <1300 playing a team >1300 – do they not get choke points in their <1300 column? Do they get their kills: deaths in the >1300 one? Does the >1300 still get them? If a 1299 plays 1301 although evenly matched are choke points applied to one and not the other as if they where a <900 squad?

“Choke” isn’t just going 6-10 it’s doing that vs a squad line you should beat, it’s also easier to choke on squads with less talented players as you will have less protection when playing against the top teams, if you took some of the top players from the best squads and put them in a borderline squad they will improve the squad but also find themselves getting beaten more heavily more often.

I know what your saying is, this would be a simple-ish to add stat that might be handy, but it is more subjective than simple K : D etc. a fairer one open to less abuse might be just to put the average rating of the squads played to earn the K : D ratio to show who is farming the ratio off the bottom squads.

[Lofty]

Seriously though, despite all the rationalization you tried to flood into here with that, the point is after 100 games, checking a Vet stats and a Newb stats, I can guarantee there will be an average formed as to how often or how much less often chokes are between the two.

Should elim no longer be ranked because you could be playing completely different types of players at different times of day and it could completely make the difficulty easier or tougher?

And that's the idea, is to give people a general idea of how consistent players are at maintaining kills above deaths, or going negative no more than -3. It's not meant to prove that it will always put out accurate results in all squad matchups or lineups etc etc, but after within the span of a bunch of games, I can guarantee the better players will find ways to stay less than -4 in most games regardless of who is on their team or who they are playing against, than worse players. And that is what these choke counters will underline. If your anal rationalization was applied to the game, we would have no stats on TWD or ranks in elim.

[Soilderz]

Aim % in DDs ^____^

[Pressure Drop]

Seriously though, despite all the rationalization you tried to flood into here with that, the point is after 100 games, checking a Vet stats and a Newb stats, I can guarantee there will be an average formed as to how often or how much less often chokes are between the two.

Should elim no longer be ranked because you could be playing completely different types of players at different times of day and it could completely make the difficulty easier or tougher?

And that's the idea, is to give people a general idea of how consistent players are at maintaining kills above deaths, or going negative no more than -3. It's not meant to prove that it will always put out accurate results in all squad matchups or lineups etc etc, but after within the span of a bunch of games, I can guarantee the better players will find ways to stay less than -4 in most games regardless of who is on their team or who they are playing against, than worse players. And that is what these choke counters will underline. If your anal rationalization was applied to the game, we would have no stats on TWD or ranks in elim.

yeah that’s exactly what i said no stats and no ranks ... but hey it's either completely against all stats or to be for meaningless stats not based on proper statistics but amazing concepts like choke count. So hey while we are at it, why not an pwnage bar and everytime you get 15+ you get 100 points added to your pwnage bar, since noobs r 2 ez if the squad is <1300 you get 200 lulz points for every 15+ killz.

if you can't take criticism or someone mildly analysing your idea then don't put it out there. Thats just my opinion on your idea. You are asking people to give up their time to code and work out methods to bring in your idea so it should be able to withstand some scrutiny.

a K : D ratio that is already implemented should give you a general idea of how well players can stay positive. You are now complaining it isn't accurate enough and we need a choke column and to separate the stats from newbs to vets

i still think that's what TWL is for, pay more significance to TWL performances than TWD and TWD stats should always be taken with a pinch of salt.

[Money]

like i said, wont work and its pretty useless, is it that hard to look at ak and ad?

[Weaver]

like i said, wont work and its pretty useless, is it that hard to look at ak and ad?

Not helpful. You speak for yourself alone when you say it won't work. I use threads like this one to generate ideas, people without that "won't work" attitude have them sometimes You should try it!

[Lofty]

And I am taking your criticism, and I am giving mine right back. There's nothing wrong with this situation, you give me your idea, I give you my idea on your idea, etc.

I agree the thing should have to be able to withstand some scrutiny, and the averages would work out pretty well in the end, considering even now the stats on TWD are worth shit as you could always subjectify it and say 'I had a bad morning that day' or 'my team was never around for some reason' or 'the lag was bad' etc etc.

[Reaver]

I think it would be nice to have something like this, the only catch would be that you would have to monitor who they're playing vs. Maybe you could set it up so that you have an Average score vs teams that have 1400+ rating, an average score vs teams that are 1200-1400, an average score for teams less than 1200. Also, there was another thread up that was talking about awards. There are a lot of games that have awards these days, halo and call of duty and I would kind of like to see something like this as well. You could start out simple with some easily detected awards, possibly something right from the TWD websites statistics. I don't wanna derail the thread but it is something to think about, you could give the awards names, have an image in the players profile on the TWD website for the awards they've received and if you click on it maybe a short explanation for how they received the award. There's plenty of ideas: 10:5 K ratio (for javs or warbird), average 2.5 reps per deaths for sharks (whatever), you can think of these easily, create an image, give them a name (Ironman, Suvivor, etc.) and then implement them into the TWD website.

[Jones]

This should be in TWL. Consistently getting good averages vs. good squads give a "general" idea about what a player can do, but it is still inaccurate, due to different lineups/4v4-format/whatnot. Surely it would be "awesome" and you might be able to track stats down easier, but stats are not always the best way to determine wether you are consistent or not.

As another proposal I would like for a "how many minutes and seconds into the game people die out." It is more useful when I'm checking stats for my squad and see some people go 7-10 but maybe play their last 3 deaths alone vs. 4.