Jump to content
Garbad

Chicken or the Egg: Does winning allow more time/opportunity for good stats or do good stats create wins?

Recommended Posts

Does winning allow more time/opportunity for good stats or do good stats create wins?

 

Good stats = Good play, you get good stats before you get wins.

Playing well means you get more wins. You dont get wins, then play well..

Link to post
Share on other sites

Is there perhaps a law of diminishing returns, a Wn8 where it essentially doesn't matter how much higher it gets as the win is already mostly guaranteed?  Those massive Wn8 games tend to be more about rushing around in garbage time to get the last of the damage vice very close nail biters where your last shot is the winner. 

Link to post
Share on other sites

There certainly is. Poltto solos at 69/70% WR with 6k WN8. Another player could probably solo at 66-67% with less than 4k. After a while you need a lot of WN8 to make WR gains.

Link to post
Share on other sites

There certainly is. Poltto solos at 69/70% WR with 6k WN8. Another player could probably solo at 66-67% with less than 4k. After a while you need a lot of WN8 to make WR gains.

Or more pointedly, why did I win more games at @3200 dpg in my T-62A than he does at 4500 dpg?

Link to post
Share on other sites

 

 

2.  I play below my average ~2/3rds of the time.  I suppose this means I'm not very consistent, yet I don't think consistency matters.

 

 

Could you explain this a bit more, I'm curious.

 

I think it's very similar to how me and gandaran play. We platoon constantly and have similar play styles with 2 very different end results. Take our 215b's for example he doesnt have the mastery badge, but yet he is very consistent with his dpg. I on the other hand do and have lower dpg because I swing back and forth from good games to bad because I am a huge risk taker in pub matches (i.e. putting pressure at every waking moment on the enemy to keep them panicked) whilst gandaran is more opportunistic than I. The combination of the 2 still at the end of a given session gives the same W/R and very similar wn8.

 

TLDR Good games will out do the bad if you're more inconsistent than other people.

 

Also when I watch your stream the passiveness in your play pisses me off, but yet you do yield result. I guess it's because I just yolo in heavies because I can.

Link to post
Share on other sites

To use a more practical example, assume someone is taking some random course in school that shall be named "X." Generally speaking, most people are satisfied with a ninety as their "A." Let's assume that an "A" reflects as a unicum in World of Tanks, just for comparative and analogical purposes. Assume that the aforementioned person has ten grades that make up their average. The "consistent" way to get that ninety would be just to have ten nineties. However, there's multiple ways to be "inconsistent" and still get ten nineties. For example, take the following grade set:

 

90

100

95

98

70

80

70

100

100

97

 

That averages out to an exact ninety. It's wildly inconsistent, yet it still achieves the exact score as a "consistent" one. However, this tends to break out better in World of Tanks due to the "limits" of WN8. Generally speaking, a zero WN8 game will bring down your average far less than a ten thousand WN8 game will bring it up, assuming your average is sub-five thousand. Thus, WN8 lends itself to averages, just like a grading system, so "inconsistency" can easily be made up for by the occasional kick-ass game.

 

This method of 'consistency' is probably quite common amongst the green-blue level, where consistently performing at the upper boundary of your skill is very difficult, certainly from my experience. At the moment, I'm usually playing sets of 15 games (5 each in 3 tier 9 tanks that I'm grinding, the Cent 7/1, E50 and E75). Typically recently, I'd say of the 15 games, 8-10 are in the 1000-1700 WN8 range, 1-3 in the total tomato range (sub 1000 WN8), and the rest, being around 2-4 games, is over 1700 WN8. In a number of those cases, those games are well over that number (2k+ WN8 range).

 

That tends to give you overall/recent WN8 in the upper green/blue range, particularly if the tanks that you are playing tend to be those with low expected values and thus have fairly low damage thresholds in order to hit high WN8 targets, as that makes getting those super-high WN8 outlier games easier.

Link to post
Share on other sites

I´m pretty sure good stats don´t create wins (not to the degree chasing best possible stats create losses) until You deal 10k+ dmg a game yourself. And I´m really afraid that wins do NOT allow for better stats cause it usually implies your team is capable of achieving something on its own, chewing your possible dmg.

 

I think You´d have to evaluate each game by what you did there - like area denial, scout elimination, flank breakthrough, being a bitch causing enemy flank turn their turrets at you, early cap etc (You can achieve all this without breaking purple wn8 and win games) and then separate only tactics that allow for actual dmg farming.

 

Or use XP as indicator

Link to post
Share on other sites

Or more pointedly, why did I win more games at @3200 dpg in my T-62A than he does at 4500 dpg?

 

Do you? If that's over a decent sample size and you're soloing like he is, then I don't know what to say.

Link to post
Share on other sites

Is there perhaps a law of diminishing returns, a Wn8 where it essentially doesn't matter how much higher it gets as the win is already mostly guaranteed?  Those massive Wn8 games tend to be more about rushing around in garbage time to get the last of the damage vice very close nail biters where your last shot is the winner. 

 

I think in the most common "standard" game you have campy but capable pubbies that can contribute significantly if you just nudge them forward a little. So you don't need to play a 6k game at all, because once you've dealt a kill while taking some aggro and doing 2k damage in the right place/time, you'll have ensured a breakthrough on your side of the map and barring a catastrophe - or a purple on the enemy team - your pubbies will be able to push to victory. So you could literally AFK and walk away with blue WN and the same victory. But of course we all press the attack and snap up as much damage as possible... It's certainly more fun.

 

This obviously doesn't apply to hard carry situations where you have to do everything to turn a loss into a win, but those are actually quite rare. The above scenario is far more common.

Link to post
Share on other sites

I think performance influences winning to a large degree. I do not, however, think WN8 performance is the truest measure of that performance. I think that explains why Garbad sees consistent winning across all WN8 levels.

I do not like WN8 as a performance measuring tool due to the fact that you cannot quantify certain performances in the game. One prime example is delay time. Enemy team is pushing a flank and you manage to stop the push. You may only deal 1k damage, but you manage to keep them in place for 30 seconds or a minute. That's enough time for help to arrive, but after your death, or time enough for your team to cap thus securing a win. You didn't perform well according to WN8, but your delay may have been the difference between the enemy capping you out.

 

Obviously, when you perform at the 6k+ level wins tend to go up with that. Those carry games are just that, victories carried on your shoulders. But for the higher level skill play I think you have more influence on a win then you realize, even if you managed to contribute very little statistically.

Garbad would be one of the best examples of a player who immediately contributes just by being in a match. The enemy team will likely make poor plays to kill him. They might be hesitant to push a flank he is sitting on, artillery will target hill over higher value targets (I am sure there are times where Garbad is not the most important person to shoot with arty), and other situations where pubbies make bad choices to kill the infamous super purple.

You can't measure those things, but I guarantee you that your mere presence on the battlefield can swing the fight in your favor without ever firing a shot.

Link to post
Share on other sites

 

I'm not a great player - I just like looking at stats.

So judging by this - what I'm thinking is that win rate and wn8 can be also influenced heavily by vehicle selection?

 

Thoughts on this?

 

Stats obtained from vbaddict.net

 

 

Yes, both can be influenced by vehicle selection.  The old Hellcat was ridiculously good at carring games, boosting our WR.  Even at Tier 8 against a better team, if I was in a platoon of pre-nerf Kitties, I liked my odds.  And damage expectations change by tank (e.g., E75 vs. IS-8) and by class (light tanks are notoriously good Wn8 padders if you're good in them).  I don't think many purples would deny that their choice of tanks is part of the reason for their stats.  Now these aren't always simple rules like the Hellcat, as sometimes a tank class' ability to carry is related to skill.  So for example, British armored TDs have good WRs, but they are generally considered poor choices for Wn8 and really high WRs because they aren't flexible.  Or, as I mentioned, light tanks are favorite purple padders.  But you have to be good to pad in them.  The average player would just YOLO and die.  So while tank selection matters, it also intersects with your individual abilities and aptitudes.

 

Or more pointedly, why did I win more games at @3200 dpg in my T-62A than he does at 4500 dpg?

 

I think there are two potential reasons here.  First, and most obviously, you might be doing damage that isn't helping your team.  Running around and killing arty while they cap.  Farming a camping KV-4 versus a T-54 that's murdering your team.  Your sample sizes are small enough for those things to matter.

 

Second, and this is more mathematical, you have a truncated deviation of statistics.  You never really hit low WRs.  I think you'd find that Wn8 is a better predictor of winning if you had a wider distribution of scores, including very low WRs and Wn8s.  In other words, a server-wide comparison should reveal a stronger correlation than a within-individual correlation.  So compared to the average player, your Wn8 is a good predictor of winning.  Compared to yourself, it's less so.

 

For those who have the excel data, could you PLEASE run a simple pearson correlation?  There's an automatic function for it in Excel, so it's easy to do.  Select the stats tool, CORREL function, and then just highlight your Wn8 colum and W/L column.  Ideally it should be a Spearman's correlation, but this will hopefully give some indication of W/L relationships with Wn8.

 

Put differently, if you were to include Wn8 in combination ith a bunch of other factors (motivation, reflexes, map knowledge, strategic knowledge, etc.) it would probably be a fairly weak predictor for any one individual as you can clearly win games without doing a lot of damage.  But overall, at the population level, it remains an important predictor because it not only predicts a useful skill (dealing damage) but also a host of other skills that most high Wn8 players have (e.g., motivation, map knowledge, strategic decision making, etc.).

 

It's a good question and I'm glad you asked it Garbad.  If someone can run actual stats on it, that would be even better.

Link to post
Share on other sites

2.  I play below my average ~2/3rds of the time.  I suppose this means I'm not very consistent, yet I don't think consistency matters.

 

It would be surprising to find strong players for whom that was not true, and it isn't about consistency.

 

If you played 9 games at 3000 wn8, then 1 game at 4000 wn8, you'd have 90% of your games are below your session average, even though you were consistent and played well.

 

The problem is that you can only go so far below your average on a crap game (you are still positive wn8 on a zero damage loss), but the upper limit is extreme.  You can have monster games that drive up your average more than you can drag it down with bad play.  

Link to post
Share on other sites

And that's why averages aren't very useful when dealing with statistics, the median being a more insightful metric.

Link to post
Share on other sites

To that end, I've noticed when asked to produce replays of my "average" games, it was nearly impossible to find any, since the average is composed mostly of high damage games and a few low damage derps intermixed. The derps drag the average down significantly without affecting the median very much.

Link to post
Share on other sites

Or more pointedly, why did I win more games at @3200 dpg in my T-62A than he does at 4500 dpg?

 

My guess is you triple unicum platoon a lot whereas he is almost 100% solo.  Sample size is probably also an issue.

 

I don't even get your rainbow chart on the front page.  You have less than 350 games listed, and in each colour category a single game is worth around 3%.  How do you expect to do any meaningful statistics on that data?

Link to post
Share on other sites

My guess is you triple unicum platoon a lot whereas he is almost 100% solo.  Sample size is probably also an issue.

 

I don't even get your rainbow chart on the front page.  You have less than 350 games listed, and in each colour category a single game is worth around 3%.  How do you expect to do any meaningful statistics on that data?

It was a solopub challenge.

Link to post
Share on other sites

Unless there's been another, it was 100@72%. Poltto has 356@70%. I think Garbad's sample size is like 150 games too small for me to take that at face value, but we can still infer that Garbad solo'd it in the same neighborhood as Poltto despite not having nearly the personal results.

Link to post
Share on other sites

Unless there's been another, it was 100@72%. Poltto has 356@70%. I think Garbad's sample size is like 150 games too small for me to take that at face value, but we can still infer that Garbad solo'd it in the same neighborhood as Poltto despite not having nearly the personal results.

Specific tank, not so much.  But 2500 games @ 70% vs. pollo or other solopub wn8stars, yes.  Most wn8 stars are a lot closer to 65% solopub.  Most people with wn8 as shitty as mine are more like 60% shitlers.

Link to post
Share on other sites

The answer is that if your goal is to win, dealing damage is *a* means to an end, not the only means.

 

What's happened on this site over the last 18 months is that winning is seen as random luck/divine providence and dealing damage is considered the actual end goal.

Link to post
Share on other sites

I thought we (well mostly Crab, meant it like the collective Wotlabs) already more or less answered a lot of these questions about how noisy wn8 gets after purple.  I'm no mathematician, but if your wn8 increases by >1k and you only get like 2-3% more wins... mysterious as fuck.

Link to post
Share on other sites

“Talent hits a target no one else can hit; Genius hits a target no one else can see.”

-Arthur Schopenhauer

 

We've raised a generation of players who think the peak of skill is clicking down a corridor 10% more often.

Link to post
Share on other sites

The answer is that if your goal is to win, dealing damage is *a* means to an end, not the only means.

 

What's happened on this site over the last 18 months is that winning is seen as random luck/divine providence and dealing damage is considered the actual end goal.

I remember seeing Valachio's review of the Conqueror and the one thing that stuck out in my mind was him chasing high health tanks to get more damage in rather than finishing dangerous, low-health tanks off.

 

Just can't abide.

Link to post
Share on other sites

I remember seeing Valachio's review of the Conqueror and the one thing that stuck out in my mind was him chasing high health tanks to get more damage in rather than finishing dangerous, low-health tanks off.

 

Just can't abide.

I watched polotto do that n his replays too.  I saw him pass on killing a reloading waffle to shoot at an afk lowe.  The waffle clipped up and raped two of his teammates.  That was so cancerous I stopped watching.

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...