Wednesday, June 1, 2011

The Virtue of Numbers

If you are a techie or a geek or (perhaps) a far-sighted visionary this post is for you. As some of you will likely know Virtue has been developing a chip that piggybacks your gun's board and collects data it reports remotely. The NPPL have been trying it out the last two events to track the pro teams' markers and make sure they aren't exceeding the 15 bps cap (that is a new rule this season.) The result, so far, is that Virtue can identify both when and how often a chipped gun shoots. This means, in a very basic way, they are collecting data on shots fired, by however many active players (and guns) during a defined game period. For starters this means they have no trouble at all enforcing the cap.
In the meantime Virtue is also trying to devise different ways of looking at their accumulated data to discover if there aren't numerous possible combinations and thus associated interpretations of that data that might be useful to teams and players. Given that they are roaming in uncharted territory all the distinctions and combinations require identifying and naming--along with considerations of their potential utility. For example, the chip is called ROF-fi (raw-fye). And currently the data gathered is simply stats 1, stats 2, etc. Stat 1 is the amount of time a player spends shooting vs. not shooting over the entire game period. Stat 2 is the amount of time a player spends shooting vs. not shooting during the period of time they are actively playing the game. Which leads directly to Stat 3 which identifies the point in a game in which a player stops shooting and doesn't shoot again--thus approximating when they were (likely) eliminated--although I suppose it's possible a player might not actually be eliminated in some rare circumstances and simply doesn't discharge his gun again. Keep in mind all this data and more is being accumulated on (in the 7-man format) 14 players simultaneously which creates a fairly enormous potential complexity.
A couple of simple stats the Virtue nerds passed along in real time at the Chicago event was the raw volume of paint teams shot off the break and the raw totals of paintballs shot in a game. The OTB number was of greater interest to me as it would be a very unusual game where the winning team didn't shoot more paint than the losers. (Think about it. D'oh!)
Chris from Virtue was kind enough to both express an interest in my opinion (and any follow-up ideas) and permit me to post about the on-going process of building unique value into the chip. I thought some of VFTD's regulars would find all this interesting and if you have any ideas, thoughts or suggestions you'd like to offer feel free to do so in comments. If that is insufficient I can't stop you from contacting Christian Williams at Virtue directly. (Sorry, Chris. Just in case you get inundated.)
To get you started thinking about the possibilities here's a suggestion I made; the Plus/Minus stat. Over the course of an event or a season it would be possible to determine an individual player's plus/minus [relative to other players] rating based on eliminations (and when that player was eliminated relative to all other eliminations during a game or point). If you took the plus/minus and compared it to the average time played as a ratio of time played: maximum playable time you could make a number of valuable if rough determinations about individual players effectiveness and that of the lines or units they played with.


Missy Q said...

I think a 'shots per kill' ratio would be pretty cool to see, both as a team stat and an individual one.

Mike said...

I agree with you Baca.

+/- is fairly obvious - in paintball this is a big one. Points won vs points lost. Or like Call of Duty, Kill/Death ratio.

Missy ! - "shots per kill" ratio would also be neat however less telling IMO. In theory wouldn't a back player have a larger ratio then a front player? May be a slight generalization but who knows...

And here's a question - when can we start seeing some data?!

houdini said...

It would be great if they could devise a way of tracking the number of shots based on a time scale from '0' minutes to the end of the game so that rate of fire could be charted against clock time.

This would give a fair idea of rate of shots off the break per player and to some extent how much paint certain players use to lock down lanes or contain players in bunkers, how much they use in one on one situations or bunkering situations and even how much their rate of fire changes during reloading situations.

(Of course the graph would need to be compared side by side with video or players recollections of their game)

You could then break down these stats and compare for example, paint usage in the first 10 seconds of a game with games won and games lost. This could be used to determine to some extend the accuracy of a team off the break or improvements in laning accuracy from tournament to tournament (If someone was also actually charting stats for kills off the break in the first 10 seconds)

chris-virtue said...

Hi Houdini
We actually have that now. Its interesting to see the volume of paint as the game progresses, and at what point the teams trade places in terms of volume shot and whether they closely follow each other or at what point one team pulls away.
Thanks for the post Baca and other for the thoughts. Plus minus composite is a great idea. I would like to get a lot of data and charts out there, but we have to get a nice online database with charting and comparisons set up. Still looking for a few good men on that front. If anyone knows some good nerds, we.d love to collaborate. Keeping in mind we'really already spending more than we can afford on this just because it's really cool...

chris-virtue said...

Damn tablet keyboard again... we may make electronics but Im a Newb with phone-tablet typing.