Jump to content


Photo

Sabermetrics In Nerf

Baseball's stat tradition + our hobby

74 replies to this topic

#1 Evil

Evil

    Fucking Copout

  • Members
  • 1,156 posts

Posted 15 April 2007 - 02:43 PM

From Wikipedia:

Sabermetrics is the analysis of baseball through objective evidence, especially baseball statistics. The term is derived from the acronym SABR, which stands for the Society for American Baseball Research. It was coined by Bill James, who was among its first proponents and has long been its most prominent advocate known to the general public.

From David Grabiner's Sabermetric Manifesto:

Bill James defined sabermetrics as "the search for objective knowledge about baseball." Thus, sabermetrics attempts to answer objective questions about baseball, such as "which player on the Red Sox contributed the most to the team's offense?" or "How many home runs will Ken Griffey, Jr. hit next year?" It cannot deal with the subjective judgments which are also important to the game, such as "Who is your favorite player?"


I'd like to dedicate a discussion to the possibility of "Sabermetrics" being a part of Nerf. I'm a huge baseball fan and its true that baseball is the only sport where statistics tell real stories and everything on the field is recorded. Because of this, patterns develop and history becomes part of the statistics; Roger Maris and 61 home runs, Rickey Henderson and 130 stolen bases, Nolan Ryan and 5,000+ career strikeouts, etc.

I'm NOT suggesting we change Nerf or play differently, and I'm NOT proposing that Nerf become more organized like a league or emphasize more of a competitive nature, but I think that good things can come of discussion and this seems to be an interesting topic no one has covered yet.

Don't bother posting "This would ruin Nerf". Thats not the point of this discussion.


Do you guys think their could be definitive statistics that could be introduced to Nerf? Even if they are limited in usefulness? What stats would be most effective for gauging a player and his/her production? Lets get creative.

Important baseball stats include batting average, on-base percentage, strikeouts per 9 innings, earned run average and of course home runs, doubles, triples, stolen bases, etc. Here is what I've come up with so far:

---
KEY:

KILLS - HOW MANY TIMES A PLAYER HAS SUCCESSFULLY HIT ANOTHER PLAYER.
HITS - HOW MANY TIMES A PLAYER HAS BEEN HIT HIMSELF/HERSELF.
ASSISTS - OCCURS WHEN A TEAMMATE'S DIRECT ACTIONS PRODUCE A KILL FOR A TEAMMATE. AN ASSIST MUST BE ACKNOWLEDGED BY THE PLAYER WHO GOT THE KILL, NOT THE PLAYER WHO HELPED. THIS STATISTIC IS ATTRIBUTED TO THE PLAYER WHO HELPED.
ROUND - ONE UNIT OF PLAY. ONE GAME.



KA (kills per round or kills average): Essentially a Nerf batting average or a pitcher's strike out/9 innings (K/9) average. Divide KILLS by ROUNDS.

Example: One Man Clan plays 6 ROUNDS in one day. He throws down 41 KILLS. His KA stands at 6.83 per ROUND.

HA (hits per round or hits average): Designates how many times, on average, a player is HIT during a round. Divide HITS per ROUNDS.

Example: Rambo plays 6 rounds in one day. He is hit 13 times. His HA is a mediocre 2.16. Meaning that if he is playing 3/15 rules he was hit just over twice a round.

KHD (Kills/Hits differential: KILLS divided by HITS. Most effective as a career statistic with a large sample size.

Example: Vacc compiles 88 KILLS with 45 HITS. His KHD is 1.95. For everytime he is hit, he hits an average of almost 2 players.

TIQ (Total Influence Quotient): KILLS + ASSISTS divided by HITS. Perhaps the best statistic to gauge how much a player contributes in one round or many.

Example: Evil compiles 7 KILLS, 2 ASSISTS, and 3 HITS. His TIQ is 3.00. For every time he is hit, he has helped to produce 3 KILLS.

W% (Win percentage): If the statistical sample was large enough (lets say 400 rounds +) this would indicate how much a player contributes to winning if his stats are considerable. Also a good stat because it accounts for several variables that might not be expressed in conventional stats. Divide WINS by ROUNDS.

Example: DarkShrimp plays 600 ROUNDS in one year. He WINS 388 of those. His W% stands at 0.64 meaning that he WINS approximately 64% of the time.

Edited by Evil, 19 April 2007 - 01:16 PM.

  • 0
2007 Great American GoreFest Champion (Aug. 4, Apoc)

#2 warx

warx

    Member

  • Members
  • 53 posts
  • Location:Crookston, MN

Posted 15 April 2007 - 02:58 PM

It would definitely depend on who took the statistics. If the players themselves kept track of their own, inflation would occur and we'd soon have little Jimmy averaging 47 kills with -2 deaths per game.

Perhaps a rotating, "designated stat keeper" policy could be introduced.
  • 0
The perfect signature does not exsist.

#3 Evil

Evil

    Fucking Copout

  • Members
  • 1,156 posts

Posted 15 April 2007 - 03:20 PM

It would definitely depend on who took the statistics. If the players themselves kept track of their own, inflation would occur and we'd soon have little Jimmy averaging 47 kills with -2 deaths per game.

Perhaps a rotating, "designated stat keeper" policy could be introduced.


An interesting prospect definitely, to have referees for matches. Might not be that difficult to do if you had lots of Nerfers and a handful who were willing to sit out during one round a day and do it. It would certainly help with called hits and whatnot. My only issue with this is that a Nerf field is so spread out that it would require players to honestly compile or remember their own statistics accurately or at least accurately enough. So this process would be far from flawless.

Also - I just added some easy statistics that could be introduced to Nerf but one important one I could not quantify; ASSISTS. Any ideas?
  • 0
2007 Great American GoreFest Champion (Aug. 4, Apoc)

#4 Flaming Hilt

Flaming Hilt

    Member

  • Members
  • 1,070 posts
  • Location:Tacoma, Washington
  • State:Washington
  • Country:United States

Posted 15 April 2007 - 03:30 PM

An interesting idea but one thing to take into consideration is round type and number of players involved. For example, in baseball, there's always the same number of players doing the same thing. In Nerf, however, you have a different number of players playing different games types every round.

IE, I play with my clan a lot which usually ends up being 5-7 people in a team battle. The point of team battle is kills, and we usually play 2-3 hit kills, so of course with 5 people a round and 3 hit kills I'm going to have a better average than someone who plays 1v1 and 1 hit kills. They could only accumulate 1 kill per round whereas I can get up to 9. So regardless of the fact that I may be not as good a player, my stats look better because I play with more people and use a different game type.

Another example of game type affecting stats would be CTF. This isn't the best example but it works. In CTF, the point of the game is to capture the flag (duh), so of course you're going to worry more about the players with your flag or in your way if you have the flag than anyone else, and you may conserve your shots until absolutely needed. Whereas in a deathmatch, the point is not to capture a flag but to shoot the other players, so of course you're going to have more kills per round (kpr? KA, as you called it).

Just something to improve upon, I guess...
  • 0
" Beneath this mask there is more than flesh, Mr. Creedy. Beneath this mask there is an idea...
...and ideas are bulletproof. "
V

#5 Forsaken angel24

Forsaken angel24

    Member

  • Contributors
  • 2,471 posts

Posted 15 April 2007 - 03:32 PM

I like the idea but would have to agree that this, much like ranges can be exagerated(sp?) and lied about.

We keep a tally for the assassin games we do.
How may kills you gotten.
How many assists you have.
How many headshots you have given.
How many times you have died.
How many games you have played.
What kind of awards have you received.

All of these things are kept recorded and carry on into the following games.
I will explain more on these in my article/book on Assassins coming soon.
I do like how you guys would keep score though. But how could you keep it legit?
Maybe like some kind of scoreboard at the end of every round when everyone can see and agree on whether they were hit or not at the end of the game. Kind of like a laser quest sheet at the end, listing who shot who etc.
  • 0
I don't get my kicks out of you,
I don't feel the way I used to do.
I know its bad,
After what we had,
But I’m just not the angel you knew.

#6 Evil

Evil

    Fucking Copout

  • Members
  • 1,156 posts

Posted 15 April 2007 - 03:38 PM

The east coast plays almost exclusively 3/15 rules in a team deathmatch format. I'm not sure how much the number of players would influence the score and I think that even if that number were to vary, it would only influence the stats so much because a good player is going to do well regardless of how many players he/she is playing against. This skill would be reflected in KA, HA and KHD.

Again, much like calling hits, tallying them would require the complete honesty of the players themselves.

It is easy to recall how many times you have been hit during a round because at max it would be 3 times for us east coasters, but tallying hits would require a disciplined effort in more intense rounds.

I'm thinking that at my next event I will try and tally some numbers and see how effective it is at gauging a single day's play. Just as an experiment to share with some of you who seem interested.

Now that we've gotten past the legitimacy issues and recognize the flaws inherent to what we're trying to do, consider 3/15 rules to be the standard and think of what other stats might be useful. Get creative. I know we have some smart people here.

Edited by Evil, 15 April 2007 - 03:39 PM.

  • 0
2007 Great American GoreFest Champion (Aug. 4, Apoc)

#7 Groove

Groove

    Certified Badass

  • Founders
  • 1,673 posts
  • Fucks Given:0
  • Location:Irvine
  • State:California
  • Country:United States

Posted 15 April 2007 - 04:28 PM

I think that it's an interesting idea, something that could be looked in to further...

I think Flaming Hilt has a good point regarding game size. The larger games will logically give you more opportunities to score hits, as well as be hit, because of the number of people playing. Here's my suggestion:

A player's stat listing could be comprised of three or more categories of stat listings, based on the number of people present. The first bracket could be games that are from 2-10 participants, the next one being 11-20 participants, then 20+ participants. Each bracket would have the respected ratios and statistics that Evil suggested. Then, perhaps the averages from those 3 or so categories could somehow be translated into the overall statistics for said player.

Awards can be given on the basis of achieving a certain ratio, score, or hit with a certain weapon, or style of gameplay. I don't know specifically what the awards could be, perhaps they could be given clever names of some sort...Most Headshots in a given round, Barrel Taps, Consecuative hits, Most Aggressive / Defensive, Sharpshooter, etc...I'm sure someone who plays games like Halo 2 could offer some better insight and suggestions here.

Again, the only problem is making it legit. A referee position, while definitely a decent idea, is not practical in larger games because he/she can't see everywhere at once. And Nerf is not like Lazer Tag in the sense where guns and vests can feed hit statistics electronically. The Honor System only works if you Honor the System. That is the biggest obstacle I see as of right now, and I don't have any real ideas to put out there regarding that.


-Groove
  • 0

"Too close for missiles, I'm switchin' to guns"


#8 CaptainSlug

CaptainSlug

    Resident Mad Scientist

  • Administrators
  • 4,761 posts

Posted 15 April 2007 - 05:48 PM

As far as tracking in a 3/15 deathmatch goes each player wouldn't have to keep track of who that person shot, but who it was that shot them. Then write all of that down at the end of the round.
That way you can't really inflate your own numbers because it's only the "victim" that's reporting. You can then draw a Hit/Kill ratio from the reported totals.

I personally don't really see the point of tracking such statistics though.

Edited by CaptainSlug, 15 April 2007 - 08:02 PM.

  • 0
The little critters of nature, they don't know that they're ugly. That's very funny, a fly marrying a bumble bee. I told you I'd shoot, but you didn't believe me. Why didn't you believe me?

#9 One Man Clan

One Man Clan

    TOFTS

  • Contributors
  • 2,170 posts
  • State:New Jersey
  • Country:United States

Posted 15 April 2007 - 07:06 PM

This would ruin nerf. Keep your hypothetical shit on the "other" forums, queer.
  • 0
I hate you.

#10 Cmdrmack

Cmdrmack

    Member

  • Members
  • 545 posts
  • Location:Decatur, GA
  • State:Georgia
  • Country:United States

Posted 15 April 2007 - 07:59 PM

I like it, but what about an international scoreboard type thing? You can see how you rank next to other players. Only problem is, some people don't have the same kind of competition in their area. I might rule the roost at Presbyterian College, but against the NJers I might get slaughtered. Or vice-versa, I might suck down here, but rule up there.

Still, that'd be fun stuff.
  • 0
QUOTE(Predalien_Ro @ Apr 7 2008, 10:24 PM) View Post

Oompa: FECES!? Who in their right mind would try that shit!?


Bigger Salvo

{FWPC}

#11 Evil

Evil

    Fucking Copout

  • Members
  • 1,156 posts

Posted 17 April 2007 - 01:36 PM

Do you guys think it would be possible to numerically gauge an assist? This is a big statistic in sports such as hockey and soccer and I know that in Nerf that my teammates (through their actions) have certainly made it possible for me to get a kill and I'm certain I've done the same for them.

In other sports an assist is easy to record. Player A pass to player B who shoots and scores. Player B gets the goal. Player A gets the assist.

This might be a subjective statistic for a war-game like Nerf, but what do you guys think?
  • 0
2007 Great American GoreFest Champion (Aug. 4, Apoc)

#12 Cmdrmack

Cmdrmack

    Member

  • Members
  • 545 posts
  • Location:Decatur, GA
  • State:Georgia
  • Country:United States

Posted 17 April 2007 - 02:07 PM

Interesting idea, but I think the person who scored the kill would have to acknowlege the assist, I think that'd be the best way to do it. Sort of an "I couldn't have done it without..." type thing.
  • 0
QUOTE(Predalien_Ro @ Apr 7 2008, 10:24 PM) View Post

Oompa: FECES!? Who in their right mind would try that shit!?


Bigger Salvo

{FWPC}

#13 nerfer34

nerfer34

    Member

  • Members
  • 1,594 posts

Posted 19 April 2007 - 10:55 AM

I've been doing this for the past year. We ALWAYS play for competition, but we don't take it too seriously.

It is fun to say who is the best overall and who is better at certain things. But there is a lot of cheating and lying, you really have to play with honest people for this system to work.

We write our stats down on paper, so far we have 7 sheets of paper down. We sometimes forget to record....

Our group keeps track of- headshots, kills, deaths.

1 headshot kills or 2 body shots kill.
  • 0
<!--quoteo--><div class='quotetop'>QUOTE</div><div class='quotemain'><!--quotec-->God Damn it Groove, you stole my kill.<!--QuoteEnd--></div><!--QuoteEEnd-->-OMC

#14 Evil

Evil

    Fucking Copout

  • Members
  • 1,156 posts

Posted 19 April 2007 - 01:08 PM

Interesting idea, but I think the person who scored the kill would have to acknowlege the assist, I think that'd be the best way to do it. Sort of an "I couldn't have done it without..." type thing.


That's a great idea. I'm going to include that in the first post. I know the possible upside of this kind of analysis is limited but it could help people find the holes in their game and thus become better at the hobby.

Edited by Evil, 19 April 2007 - 01:17 PM.

  • 0
2007 Great American GoreFest Champion (Aug. 4, Apoc)

#15 Rambo

Rambo

    Fear the Robble...

  • Members
  • 1,807 posts
  • Location:Allentown, PA

Posted 19 April 2007 - 02:38 PM

I suppose you could group kills and headshots differenty...I mean, I never really aim for the head, but if you really want to break it down, you could...
  • 0

#16 Evil

Evil

    Fucking Copout

  • Members
  • 1,156 posts

Posted 19 April 2007 - 04:16 PM

I suppose you could group kills and headshots differenty...I mean, I never really aim for the head, but if you really want to break it down, you could...


I would otherwise but not how we play. Think about it. A kill is a kill for us regardless of where on the body you hit.

Edited by Evil, 15 May 2007 - 10:29 AM.

  • 0
2007 Great American GoreFest Champion (Aug. 4, Apoc)

#17 Evil

Evil

    Fucking Copout

  • Members
  • 1,156 posts

Posted 17 June 2007 - 01:54 PM

At the most recent DCNO, some Nerfers volunteered to tally their kills and hits after Rambo suggested that we attempt to record our stats. With his help we've actually been able to accurately tally these stats and I thought some of you might be interested in seeing what came of them. Remember, while this might be interesting as a snapshot of one event, what is going to be cooler later on (if we keep doing this) is when we see patterns in the statistics later. One event doesn't really define a player.

Check out the raw date here: Rambo's page. And also check the first post of this topic for more info.

Evil:

Kills Average [KA] (per round): 5.44
Hits Average [HA] (per round): 2.00
Kills/Hits Differential [KHD]: 2.72 (kills for every hit taken).
Win Percentage [W%]: 66(%)

I encourage the rest of you guys that tallied your scores to post them here also in the same way I did.

I also noticed that at first it was hard to get used to keeping some kind of score, since for so many years we haven't kept personal scores. It got easier as the day went on though and if more people want to try this experiment at the next event I would be all for it. These aren't perfect statistics but they are accurate enough that I think they give us an idea of what goes on on the field.

I appreciate the four of you who tried this with me, and thank Rambo for suggesting it.

Edited by Evil, 17 June 2007 - 01:55 PM.

  • 0
2007 Great American GoreFest Champion (Aug. 4, Apoc)

#18 Rambo

Rambo

    Fear the Robble...

  • Members
  • 1,807 posts
  • Location:Allentown, PA

Posted 17 June 2007 - 02:31 PM

I recorded everyone's stats in my Treo on an Excel spreadsheet. This was pretty easy, but using a tablet might be just as easy.

Let me just say that while I don't mind recording a few people's stats every war, I'm not going to do everyone's. But, I will put them on a website if you record them.

I'm really looking forward to seeing where we can take this.
  • 0

#19 baseballnolan5

baseballnolan5

    Member

  • Members
  • 90 posts
  • Location:Raleigh, NC

Posted 17 June 2007 - 02:47 PM

This is a very good idea. If, at the end of every round, everyone went and recorded their stats on their own piece of paper they could easily keep track of their own stats. This would especially help for wars such as DEAL where it is invite only if you posted your "stats" in your signature. This would eliminate some one saying they have war a lot of war experience and then get an invite to DEAL and they end up completly sucking. Good idea with the spreadsheet rambo.
  • 0
QUOTE View Post

QUOTE

About how much are you holding in the pictures on the spools?

A shit ton.

QUOTE(Langley @ Aug 6 2007, 05:41 PM) View Post

The shot hit Andy squarely in the forehead. He was, in fact, on my team.


#20 Evil

Evil

    Fucking Copout

  • Members
  • 1,156 posts

Posted 17 June 2007 - 03:00 PM

If, at the end of every round, everyone went and recorded their stats on their own piece of paper they could easily keep track of their own stats. This would especially help for wars such as DEAL where it is invite only if you posted your "stats" in your signature. This would eliminate some one saying they have war a lot of war experience and then get an invite to DEAL and they end up completly sucking.


I don't think any invitation would ever be based off of any statistic. The reason DEAL is invite only is because of the fact that the East Coast and its Nerf community has gotten bloated in the last year or so, and we now have upwards of 30 people showing up at many events.

I don't think stats would be a good way of proving anything to the community if you were detached from it, considering how easily stats could be inflated or made up on these forums. The only reason I trust the numbers that have been posted here is because I saw all of these players in action, and because they saw how I played, have seen my stats accordingly, and have indicated that they are legitimate.

I think this is a good experiment for groups of Nerfers who have been playing for some time and for those who are just interested in adding something to the average war.
  • 0
2007 Great American GoreFest Champion (Aug. 4, Apoc)

#21 Cats

Cats

    Member

  • Members
  • 16 posts
  • Location:Upstate NY

Posted 18 June 2007 - 08:21 AM

In my opinion statistics apply to certain aspects of Nerf but falter in others. For example: It would be easy to gauge head shots, total kills or points scored (in capture the flag or hunted). An organized system crumbles under rating things like player endurance (very important), length of time between being shot, sneakiness (beloved quality of nerfers who only play at night) and ability to take cover and utilize the environment. However I'd like to see a system of nerf battle rating, but I don't foresee a definitive and universal sabermetric system in nerf.
  • 0

#22 CaptainSlug

CaptainSlug

    Resident Mad Scientist

  • Administrators
  • 4,761 posts

Posted 18 June 2007 - 08:50 AM

Just based on one I event at the very least it allows the players that recorded their statistic to gauge their own performance and how well they did. Further recording can allow them to compare their performance between venues. Beyond that I don't think there's much you can do with the numbers since they only paint a partial picture.

Edited by CaptainSlug, 18 June 2007 - 09:41 AM.

  • 0
The little critters of nature, they don't know that they're ugly. That's very funny, a fly marrying a bumble bee. I told you I'd shoot, but you didn't believe me. Why didn't you believe me?

#23 Evil

Evil

    Fucking Copout

  • Members
  • 1,156 posts

Posted 18 June 2007 - 09:28 AM

In my opinion statistics apply to certain aspects of Nerf but falter in others. For example: It would be easy to gauge head shots, total kills or points scored (in capture the flag or hunted). An organized system crumbles under rating things like player endurance (very important), length of time between being shot, sneakiness (beloved quality of nerfers who only play at night) and ability to take cover and utilize the environment. However I'd like to see a system of nerf battle rating, but I don't foresee a definitive and universal sabermetric system in nerf.


Obviously this system has limitations because of the simple way we play. But everything you listed that would make the "organized sytem" crumble are all intangibles that couldn't be gauged by a serious statistic anyway. Length of time between being shot? Unecessary if you consider the the strength of someone's round; if Groove goes 7-2 (like he did first round at DCNO), who cares how long his hits were spaced out by considering that he lasted the round and produced 7 kills?

Sneakiness? Completely unnecessary and too abstract to gauge with a statistic. We don't play at night, and if you need the dark to be sneaky than you are a garbage Nerfer. If you are effective, it will show up in your stats. If you happen to be sneaky than I suppose your HA would be down, and your KA would be respectable (since most "sneaky" players that I've seen are not the most aggressive).

We need a stat to dictate how effectively a player utilizes cover? What would something like this matter? All these statistics are attempting to do is give a general idea of capabilities of the player and their impact in a particular round, event, and later on we could analyze the trends that develop in these stats for lets say a year's worth of Nerfing. We're not trying to define players with these numbers. As far as I've seen almost everyone has their own style of play when it comes to this hobby.

Headshot tallies? Unecessary. Its cool in Halo 2, but because none of us aim for the head and our guns are not accurate enough to hit someone in the head on purpose, this is a useless (albeit cool) stat.

You don't see a "universal sabermetric system in nerf"? The only issue I really see with this system is the accuracy which we need to keep special attention to and the pool of players who want to utilize it. A Nerf Battle Rating, which you've suggested, might be something closer to an OPS rating in baseball which equates to:

the sum of a player's on-base percentage and slugging percentage.[1] Both the ability of a player to get on base and to hit for power, two important hitting skills, are represented, making it an effective way of measuring the offensive worth of a player. An OPS of .900 or higher in Major League Baseball puts the player in the upper echelon of offensive ability. Typically, the league leader in OPS will score near, and not necessarily below, the 1.000 mark.


If we could find a stat similar to OPS in baseball we might be onto something. The closest thing I have now is the TIQ stat.

What do you guys think?

Also, gauging assists I think will be my experiment for DEAL. I think these will be subjective to a teammate's view of a given situation and will only be credited to a player if their teammate feels it necessary. Essentially, you have to directly produce a kill for a teammate.

Edited by Evil, 18 June 2007 - 09:37 AM.

  • 0
2007 Great American GoreFest Champion (Aug. 4, Apoc)

#24 z80

z80

    Member

  • Members
  • 603 posts
  • Location:Larchmont, NY

Posted 18 June 2007 - 09:36 AM

Theres also the fact that most of those things contribute to the other stats. Sneakiness will mean that you get shot less. More shots during a round? Probably you will kill more people. They all contribute to other stats, that can be recorded.
  • 0

#25 Evil

Evil

    Fucking Copout

  • Members
  • 1,156 posts

Posted 18 June 2007 - 09:58 AM

I think all we really need is a few notes on the game-type and the rule set. 3.15 team battle is our conventional game and we should also make a note of how many players were involved in the day's event, so this way, like Hilt suggested, we can accurately gauge how well a player did.

Summer DCNO:

Game-Type: Team Skirmish
Rule Set: 3.15
Game Size: 20 (approx.)

Rambo, if you could put this info on your TeamSerenity.com site that would be great.

Edited by Evil, 18 June 2007 - 09:59 AM.

  • 0
2007 Great American GoreFest Champion (Aug. 4, Apoc)


0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users