Compustrike/Fight Metric

[SPX]

Does anyone know of any fights that Compustrike or Fight Metric just royally fucked up? I'm considering pitching an article on the subject but I need 5 fights to really back it up.

Thoughts? zY, I know you've mentioned one or two. . .

[Luke]

What do you mean by royally fucked up?

[SPX]

What do you mean by royally fucked up?

I mean by not counting takedowns, or sub attempts, or strikes. Basically fucking the numbers up.

[Luke]

What do you mean by royally fucked up?

I mean by not counting takedowns, or sub attempts, or strikes. Basically fucking the numbers up.


Ok I understand what you meant now.To answer your question though I have no idea

[zY|]

Compustrike solely counts strikes and strike attempts, and is done ringside live.

Fightmetric is done after the fact.

Fightmetric is, for all intents and purposes, awesome. It doesn't always tell the whole story, but for the most part it's pretty damn accurate. The only thing I can HOLY FUCK Hector Lombard! He just knocked out Jay Silva in 6 seconds!

[zY|]

Goddamn.

Anyways, the only thing I can think of that's weird about Fightmetric is the Penn/Edgar report. They recorded 72 landed strikes for Penn and 63 for Edgar, yet had 71 and 61, respectively, as power shots. Therefore they only had Penn landing one jab the whole fight, and Edgar only 2.

Now unless our definitions of powerstrikes are drastically different, something is up with that because that fight was jabcity.

[SPX]

Interesting.

Thanks zY.

[zY|]

Obviously there may be more, but that's all I can think of at the moment.

There may be some close decisions I don't agree with, but nothing stands out as egregious.

Compustrike is subject to more errors, solely due to the nature of it.

[Svino]

Goddamn.

Anyways, the only thing I can think of that's weird about Fightmetric is the Penn/Edgar report. They recorded 72 landed strikes for Penn and 63 for Edgar, yet had 71 and 61, respectively, as power shots. Therefore they only had Penn landing one jab the whole fight, and Edgar only 2.

Now unless our definitions of powerstrikes are drastically different, something is up with that because that fight was jabcity.

Don't quote me on this, but it seems like their definition of a power shot is a standing strike like you would see in boxing - reserving the label of non-power shots for the little GnP "rabbit" punches.

[zY|]

Goddamn.

Anyways, the only thing I can think of that's weird about Fightmetric is the Penn/Edgar report. They recorded 72 landed strikes for Penn and 63 for Edgar, yet had 71 and 61, respectively, as power shots. Therefore they only had Penn landing one jab the whole fight, and Edgar only 2.

Now unless our definitions of powerstrikes are drastically different, something is up with that because that fight was jabcity.

Don't quote me on this, but it seems like their definition of a power shot is a standing strike like you would see in boxing - reserving the label of non-power shots for the little GnP "rabbit" punches.

I thought the same thing, but then you see this.

http://www.fightmetric.com/fights/Penn-Sanchez.html

[Svino]

Well, OK. That kills my theory, unless different guys do different fights and they don't agree on the definition.

[triathlete]

Well, OK. That kills my theory, unless different guys do different fights and they don't agree on the definition.

That's an interesting point on this. We're doing an educational research study (I'm learning how to conduct research in a lab) and two researchers have to "code"... stuff. There's something called inter-rater reliability (http://en.wikipedia.org/wiki/Inter-rater_reliability I'm sure Svino knows about it). In our study, two people go through a potential result set and score certain aspects of it. They then compare the two reports to see if the two researchers 'saw' the same thing.

So, the point I'm getting at... I wonder what kind of training they do for consistency, and if they measure this between fights/judges, etc. If they do, then the differences between different guys doing different fights has been considered and mitigated. If not, then that might change the way you look at the reports.

BTW, I'm not an expert in this. Just introduced in the last few weeks, and of course the first thing I thought of was this forum. ::handshake::

3