FightLines Update

by Mike Fagan on Aug 12, 2008 11:30 PM EDT

If you've been paying attention (probably not), you might have noticed I haven't done a FightLines graph for the new middleweight or lightweight meta-rankings. Simply put, the results of the past month haven't really done much to change the overall structure of either graph. That's kind of been a similar issue with the 3 I have done this month (outside of the light heavyweight graph, really).

The biggest problem is that there just aren't enough fights within a month's time to make many noticeable changes month-to-month. As a result, I'm going to start staggering the graphs in a way where we can include more fights between intervals. I feel like doing one weight class every 3-4 weeks would work well. This gives each weight class 15-20 weeks of time to build up data.

Thanks to everyone who has supported the project so far.

0 recs | Comment 9 comments

Story-email Email Printer Print

Comments

I'm not too vocal about it,

but I really like the inclusion of FightLines here. It gives a pretty interesting way to view the relativity of all the fighters in a weight class, and I’m sure it requires a bit of work on your part to compile.

Thanks for the effort!

There is no such thing as innocence, only degrees of guilt.

by misterjonez on Aug 12, 2008 11:34 PM EDT 0 recs

I don’t like the Fightlines. They make absolutely no sense, and if they did they only prove how inconsistent wins and losses can be and how unpredictable fights are. You’re basically just too PC to put fighters in an order for rankings so you just spatter them around.

Besides, Fightlinker and Martial Farts peeled the cap of the whole incomprehensible idea anyway.

by godzillad on Aug 13, 2008 12:34 AM EDT 0 recs

I understand that the idea isn’t for everyone, but this…

“You’re basically just too PC to put fighters in an order for rankings so you just spatter them around.”

...is just outrageous.

by Mike Fagan on Aug 13, 2008 8:55 AM EDT up 0 recs

It's an awesome tool, to be deployed amongst other tools

such as W/L records, stylistic match-ups, overall experience, tools scouting, etc..

As pumped as I am about the new video game coming out next(?) year, it’s going to seriously create some turmoil amongst the fans, since it will probably end up being pretty final in its ranking system.

There is no such thing as innocence, only degrees of guilt.

by misterjonez on Aug 13, 2008 10:30 AM EDT up 0 recs

They make a lot of sense to anyone with decent analytical or reasoning skills. I love the guys at Fightlinker, but their critique was like shitting on great literature because it’s hard to read. Not the most convincing stuff ever.

by Luke Thomas on Aug 13, 2008 9:38 AM EDT up 0 recs

They may talk like they're crazy, but they're right...

I wasn’t going to take a stance on this, but I have to agree with Fightlinker. This is basically using MMA Math to determine who’s the best. It doesn’t work that way. Styles make fights, but fightlines is the epitome of MMA Math. Furthermore, the arbitrary logic to take out “circular” wins make no sense at all. The fact that there is circular lines is why fightlines doesn’t work. By taking it out, it’s like pushing a square peg into a round hole.

I have to give it credit for giving us a good graphical view of who’s beaten who (which should be it’s greatest asset), however, by taking out circular wins, it just collapsed its one real use: graphical representation of wins/losses. To use it for ranking fighters, it’s ridiculous.

I suggest put back the circular wins, and just use it as a quick graphical representation of Sherdog’s win/loss records. It makes much more sense that way.

by cyph on Aug 13, 2008 10:52 AM EDT up 0 recs

You couldn’t be more wrong. FightLines is so far removed from MMA Math it’s not even funny. I’ve seen every single FightLines that’s appeared and never once used it for purposes of MMA Math. FightLines aren’t used as predictors or to determine a linear ranking; they’re used as a snapshot in time of what the division’s landscape looks like. They aren’t there to be the be-all, end-all, but a graphic representation that adds to our existing understanding.

by Luke Thomas on Aug 13, 2008 11:27 AM EDT up 0 recs

That is the best way to put it. Its a Landscape, not rankings or a predictor of any kind. I see it as a tool to see if the best guys are fighting other top guys. The goal, and I use that term loosly, for a match maker using FightLines would be to make the graph more linear by matching up guys near the top.

by szucconi on Aug 13, 2008 5:44 PM EDT up 0 recs