Stats Discussion

[Enyo]

The current values with 2.2 scale is giving players with < 50% win rate a positive multiplier. There is no way that is correct regardless of how good their aim is.

You don’t think Leon, despite having only a 49% win rate, should have a positive multiplier? Win rate is not the best metric to be using to determine multipliers.

I get why you guys would think that’s most important, but your win rate is largely affected by the teams you’re given, not just you individually.

The current formula for determining multipliers is too simplistic and doesn’t take into account the multiple factors that make a “good” player in onslaught.

Maybe try replacing win rate in the formula with the node hour summary number instead?

[pooty]

Take a look at column M in the sheet.
However instead of 100 (perfect win pct), I used max for NHS (68 max in the data rounded up to 70) as the base.

[captainsnarf]

I get why you guys would think that’s most important, but your win rate is largely affected by the teams you’re given, not just you individually.

But your multiplier affects the team you are given.

I agree there is some synergy that is not accounted for when figuring out multipliers for each individual. I made the example before of Azazel always going for the middle node in MB, and Fragster always grabbing the Ion and being useful with it. It's accidental teamwork and they are more likely to win because of it, but it's not taken into account for each individual's multiplier. You put them on the same team on a different map and the synergy is gone.

If you rank by Node Hour summary that means you rank players like Vugluskr and Ankeedo higher than Anonymous. That's not the right metric either. Win % is the only thing that takes everything into account. Aim, strategy, communication, the team handed to you, how much you goof off, etc all end up in how often you win.

[pooty]

strategy, communication

These two are key and don't have any numerical representation, except as factor in WinPCT.

Being in the right place at the right time, even with crappy aim (not really needed in a tank usually, esp. mino) can make or break a round.

[Enyo]

I get why you guys would think that’s most important, but your win rate is largely affected by the teams you’re given, not just you individually.

If you rank by Node Hour summary that means you rank players like Vugluskr and Ankeedo higher than Anonymous. That's not the right metric either. Win % is the only thing that takes everything into account.

Sorry, I meant to say NodeHourSum + WinPCT, listed as Ranking here: https://www.omnipotents.com/utstats/onsrankings.php

Either way, even more so than top players with +1 multipliers, and as pooty and I have both said, it's likely the bottom feeder shitstains have more effect on whether a teams wins or not. A team full of 5 or 6 stains is always at a huge disadvantage vs. a team that only has 1 or 2, regardless of the top players being evenly distributed. I don't think the problem is so much with multipliers as it is with the way that the balancer works in that it never seems to balance the stains well.

I'm in agreement with Scrotey in that it's basically impossible to truly balance onslaught. There's just too many variables.

[pooty]

as it is with the way that the balancer works in that it never seems to balance the stains well.

I'm in agreement with Scrotey in that it's basically impossible to truly balance onslaught. There's just too many variables.

Balancer has a decent algorithm.... not sure how else you'd do it, other than than to make a full on separation for stains... and split them evenly to start.....

But between all the variables, and worse yet leaves/joins... no way to balance, and almost universally everyone hates in round/match reshuffling...

[pooty]

I still need to update the mulitplier to the updated formula. Hopefully I'll get to that today!

[Super Sanka]

Make sure you give phariz/beetchucker the worst rating possible

[pooty]

Just to be clear BeetChucker and Pharizwheel aren't the same player...

[pooty]

I put the old multipliers back.. the newer ones even the upscaled ones didn't seem to work as well.