Potential things to track
Fantastic idea, not just for spectators, but an important tool for coaches too. Statistics will remove a lot of the BS you have to hear from players about their kill ratio, how they got eliminated, etc.
Yes kill count is, by far, the most important stat, but I would like to see it broken further down, as follows:
1. Total kill count
2. OTB kills
3. Laning kills
4. Gunfight kills
5. Bunkering kills
6. Kills without getting eliminated
7. Trading kills
Also, I'd like to see the following other stats:
1. Total deaths
2. Deaths in a bunker
3. Deaths on the move
4. Deaths gunfighting
5. Deaths getting bunkered
6. Deaths where the player trades
7. Deaths where the player does not trade
A. Total team points (incl. draws)
B. Player points played (incl. draws)
C. Team paint consumption
D. Player paint consumption
E. Team on-field time
F. Player on-field time
G. Team on-field time per point
H. Player on-field time per point
I. OTB bunker
J. Secondary bunker (can be "none")
I know it might not all be realistic at first, but in time it will, possibly with the addition of gun chips that transmit some of the information.
http://www.pbnation.com/showthread.php? ... st74710422i know where your line of thought comes from.
i'm not saying the efforts and intentions are fail, but the aproach is definetly FUBAR from the moment mike paxson became a front player in the stats.
that thing is the main and more obvious symptom, and whay tried to explain are the probable underlying causes for it.
baseball and american football are very repetitive and controlled action sports.
the intervals of action are very short, and altough all players get to do something, the defyning individuals for each sport are but 2 or 3.
i'm not an american football fan for other than watching important games every now and then, but i know that for each play (interval of action), there is a quarterback, a possible receiver, and some guy that takles someone. sometimes there are only one quarterback and a tackler, sometimes someone just gets penalized for being a douche.
compared to rugby (pass after pass after tackle after pile up after pass...) football has very little chaos involved for there are way less defyning interactions in it.
i don't watch baseball at all, because it is bloody booring and tedious to do so. most of the time it's just some guy throwing a ball and some other dude swinging a bat. like a third of the times there is a third person involved in catching a falling ball, and some other guy passing it along, while people run s fast as they can in predefined directions.
in paintball, it is as if each player had the winning ball with them the whole game (because they do). and everything he or she does or does not alters the game to a point it becomes statistically relevant even when they die off the break.
and by my empirical experience, it seems like more often paintball players have to improvise and/or play outside of their predetermined jobs per match, than most other athletes do in their respective sports.
what i feel should be the right(ish) way to take on stats for paintball is to take as a whole many factors (instead of isolating them and analizing them without context) and provide a general idea instead of a bunch of many numbers for factors not representative at all by themselves or even if you group them with others.
what did i just say?
that maybe stats on team performance are more relevant, representative and apply/count better than individual and micro statistics on players.
like:
live players per break (for each team)
kills per break (for each team)
average point time (for each team)
average runners per break
average shooters per break
average hybrids (shooting and running) ber preak
points played
points won
average of players alive per each flag hanged.
etc etc
and for individual players we should only be taking into account stats that relate to the individuals performance alone.
and it's the statitians', analists' and general people's job to assign a determined "position" or "style of play" to each player based on what said player stats say, and not assigning this definition before before collecting data.
K/D ratio should be applied to each player.
average times shooting still per break:
average times shooting-running per break:
average times alive per point.
(if you read these 4 stats together, one can easily pinpoint what a player's general job normally is, and how good he is at it)
Stats on penalties should be a MUST though, for individual players.
and so on.
My general idea is to remove any and all stats that may create bias on the individual player's performance, in order to get better stats and therefore more value themselves.
because in paintball is OK to trade out your body, if your team benefits from it. it's also OK to die quickly, if this benefits your team.
it is OK to kill no one, if you help mantain your team on the drivers seat of the point.
it's ok to kill no one on a break, if you made it to your desired position on the field.
etc etc.
if history and empirical experience tells you "X" player is more than anything back player and basically helped define the position himself, but player X wants to have his stats collected and read as if he was a front player, you know your whole system is creating a massive bias on the very data source, and will give you unrepresentative conclussions.
http://www.pbnation.com/showthread.php? ... st74756696If you are using a straight K/D ratio, than it will most likely even out over a long day. If the fronts and mids are good, they will get most of the kills. However, they will die more often. The back guys will generally get fewer kills, but they tend to last longer.
Frankly, I feel that the straight K/D ratio is archaic. It is useful during games where you are reinserting, but that doesn't happen during tournaments. I have actually been working on a formula that I think provides a better picture of a player's abilities.
Note: The formula that includes deaths is not my preferred one, but I included it for discussion
I feel that using the amount of time a player was alive is better than using deaths. I feel that getting shot out at a bunker early in the game and trading out at the end should not be considered the same, which is what happens when you use the number of deaths.
What do you guys think?
http://www.pbnation.com/showthread.php? ... st74766674
