Tweet
Originally posted by Malladrin
avgRepLeft = cummulative repels / deaths
total reps shark has done = (deaths * 3) - cummulative repels.
They are all basically linked. So its probably already accounted for
-
cummulative repels = sum(num of repels left when shark dies)
avgRepLeft = cummulative repels / deaths
total reps shark has done = (deaths * 3) - cummulative repels.
They are all basically linked. So its probably already accounted for -
You can green reps also... I might not be the best shark out there but I certainly go for greens to get reps. So you can actually have repped more than 3 times when you die.Originally posted by Force of NatureComment
-
Yeah that was one of the reasons I suggested earlier in the game that counting the actual repels used is better, because often you can get 4-5 repels on a single life, you are obviously a little more worthy of MVP if you manage to green 20 extra repels in a game than the other sharks do.
Also it wouldn't be fair if you used 3 repels then greened one and died instantly...sdgComment
-
I'd just like to take a moment to say, "GOOD FUCKING JOB, FoN!" These MVP formulas are turning out quite nicely. By the time you're done tweaking them, I think we'll finally start to see people truly earning MVP, rather than spider after spider after spider getting it. Once more, kudos to FoN! (Kudos to everyone who's contributed to the refinement of his formulas as well, most notably Randedl and BLeeN.)jasonofabitch loves!!!!Comment
-
Jason speaks the truth. Well done Force of Nature!Comment
-
I still think it should be based off name length.Comment
-
Yeah, 0.8 is way to high for terr coefficient, but it should still be worth more than 0.1. For determining this think of a flagroom battle, would you rather have your spider kill say 2 enemy spiders or kill enemy terr. So whatever you think is better for a particular ship to kill think in terms of '3 dead enemy spids at an exact moment in battle is worth killing a terr'. So if you agree with that last statement than the coefficient for killing a terr is worth 3 times the coefficient for killing spids. so if spid coefficient is 0.05 then terr coefficient is 0.15. That I think would be a good way to determine coefficients, and by having a terr coefficient that high you will get spids to push more so they can td instead of camping like some of these spids do, far too many in my opinion.
-L3Comment
-
Abolish MVP and add stats like these at the end of the games:
Most Points, Most Points, Most Deaths, Most TKs, Most TerrKills, Most Repels Used, ETC.sdgComment
-
the most repels used isn't a good measure of a good shark. It's the quality of the rep, the timeliness of their death, kills, lack of teamkills. etc etc.
If shark is going to based on how many reps used, i'll just rep rep rep die. repeat. There is a lot more to sharking than repping.There once was a man from Nantucket.Comment
-
What's your point? I am just talking about stats, it doesn't have to be a measure of how good the player is.sdgComment
-
oh, sorry. Why are we keeping these stats? just because we can?There once was a man from Nantucket.Comment
-
Might as well. Interesting to look at, and really you will never be able to come up with a MVP system that could be easily and fairly award MVP to any ship...sdgComment
-
I also just found out that the current system is extremely flawed when it comes to shipchanges.
I was MVP for the entire game, and when a spider changed into a shark, they suddenly got like 4x the rating I had and got MVP, because of course they had a very good ratio/kills and 0 TKs and a 0 rating for "repelsLeft".
The only way I can think to fix that is to either do something with the rating when someone changes ships, or totally disqualify them from MVP if someone changes ships midgame.sdgComment
-
The formulas aren't perfect and they never will be for MVP. But we can try to make something for it. Those factors need to be changed. I only made those factors based on significance, but only with real stats I can change it to better ones.
Everyone keeps asking for MVP, but in the end its the whole team who takes the win in effort, not just one person. Ah well, just my opinion =P
Stab, i've already considered it about, switching etc.
It's just a programlogic which needs to be fixed.Last edited by BLeeN; 03-22-2004, 07:22 AM.Light is faster than sound. That is why people look bright, until you hear them.Comment
-
The current system is to be meant as a temp system until i get the time to implement the other 2 post system I mentioned before. I am currently unable to rectify or fix anything with the current system because I don't have access to the code anymore since I'm at a different place for atleast a month. The weights need to be fixed. I haven't had any time to actually see the system work for a lenghty time but by my intuition I can guess that I probably nerfed the jav + spider a bit too much. So, I have to increase the weights for those ships.Originally posted by Stabwound
The ship change problem was already well known. As I've said before I don't have the time or the code atm to rectify the situation. The easy solution is to ban ship changes from getting MVP. The other solution which I would implement in the autonomous MVP is that the rating of the previous ship ship will be added to the new rating of the current shipchange. Obviously to do this is not a trivial task. Firstly the stats have to be differentiated for every ship. For example each shipchange needs to have its own set of stats. So the rating can be seperately calculated for the time spent on each ship seperately. Also, the whole stats need to be accessable if you want the total stats of a player in a concise format. Again, I wont be able to rectify this at this time.
My goal for implementing this was to improve on the current system with the temp system. It is still flawed yet less so than the previous system. And for that you should be atleast a bit appreciative. If you still think that this isn't worth it. You can continue to ignore the MVP as you would before.Comment
Comment