Fantasy Fights

Intelligent, unique MMA analysis

UFC 135: SILVA Update

I’ve thought about doing this for a while, and now I’m actually going to do it. Since so much of my analysis revolves around the fighters’ various SILVA scores, I think it’s a good idea to comment about how each fighter’s SILVA score is affected after their fights, to help provide more complete analysis.

A full list of processed SILVA scores can be found here.

Heavyweights

Travis Browne and Rob Broughton are unchanged from their SILVA scores of 30.81 and 29.21 respectively. Simply put, SILVA expected Browne to beat Broughton, and he did, so there’s no reason to change either fighter’s estimate.

Mark Hunt’s SILVA score actually goes slightly down after beating Ben Rothwell, from 31.01 to 30.68. The reason is that Hunt’s PRIDE Shockwave 2005 win over Mirko “Cro Cop” Filipovic is no longer considered relevant by the system. That takes away Hunt’s highest-rated career win, and his SILVA score declines slightly as a result.

Ben Rothwell’s SILVA score takes a total beating, plummeting from 32.87 all the way down to 25.49. The first thing to keep in mind is that Rothwell’s total career resume is really not that impressive: his best win is easily Roy Nelson, and it goes downhill fast after that. SILVA does not like that Rothwell lost to a previously 6-7 opponent at all, even if the opponent is one with a “strength of schedule” as brutal as Mark Hunt’s.

Light-Heavyweights

Jon Jones goes up to 46.02 on the strength of beating “Rampage” Jackson, and as always, I have to note that his SILVA score would be significantly higher if the infamous Matt Hamill fight is counted as a win instead of a loss. In fact, his SILVA score in that instance would be 51.85, and place Jones behind Rashad Evans and Phil Davis, and ahead of Lyoto Machida in the light-heavyweight division. I think such a rating would be fair: Jones has looked spectacular against very good competition, but he hasn’t beaten the very best fighters in the division yet (Evans/Davis/Machida). Jones vs. Evans should be a HUGE fight.

Jackson’s SILVA score is unchanged at 37.40, which is as it should be. He took on a much better opponent and lost. There’s no reason to downgrade him because of that.

As for the earlier light-heavyweight fight, James Te Huna goes up from 29.99 to 32.35 since a loss to James Lee earlier in his career is no longer considered relevant. I think this is reasonable.

Ricardo Romero stays put at 25.00, and after two lightning-quick UFC losses, I have to wonder if his UFC career has come to an end.

Middleweights

Tim Boetsch goes up all the way to 35.12, and is now rated as the tenth best fighter in the middleweight division by SILVA. One of SILVA’s (potential) flaws is that it tends to overrate wins against opponents who have an outstanding record against largely mediocre or bad competition. Nick Ring fits the bill here: he was 12-0 entering this fight, but his only two good wins were against Riki Fukuda and James Head. This is something I’ll look into when I get around to tinkering with SILVA.

While SILVA overrates a win over an undefeated fighter with a weak “strength of schedule,” it doesn’t tend to overrate the fighter himself. Nick Ring stays put with a SILVA score of 29.20, which makes sense to me.

Welterweights

Josh Koscheck moves up to 34.95 with the win over Matt Hughes, and that’s a SILVA score that I’m comfortable with. Koscheck was probably underrated before, when his SILVA score was just 32.60, but his new SILVA score reflects the reality that Koscheck is a very good fighter who has some obvious flaws.

There were four fighters rated as “elite” by SILVA who I felt didn’t belong there because they were on the decline. Two of them fell sharply in the ratings as a result of last night’s fights. I’ll get to Takanori Gomi in a moment, but Matt Hughes goes down to a SILVA score of 33.03 as a result of losing to Josh Koscheck. The reason is that Hughes’s rAP (regressed Average Points) went below 43.00; fighters with a rAP of at least 43.00 are considered to be “elite” fighters, and are rated above their contemporaries. A rAP of 43.00 or higher suggests that a fighter has taken on quality competition and won on a consistent basis. For legends such as Hughes, Gomi, Fedor Emelianenko, and Antonio Rodrigo Nogueira, they were correctly rated as “elite” fighters at one point in their respective careers by consistently beating tough competition. But because each of them had so many fights, when they started to decline, each loss didn’t hurt their rAP enough to take them out of the “elite” tier. Hughes and Gomi are finally out, but they really should have been out before last night.

Lightweights

And now we get to Gomi, who suffers a much worse drop than Hughes did. In fact, Gomi goes all the way down to a SILVA score of 27.82, as he’s hurt by a loss to Sergey Golyaev on his record, and hasn’t beaten any great opponents anytime in the recent past. Gomi is clearly done as a relevant fighter in MMA, and it’s appropriate that he’s no longer rated as an “elite” fighter, but like Hughes, that should have happened before his loss to Nate Diaz.

For Diaz’s efforts, he goes up from a SILVA score of 29.21 to 31.91 due to beating Gomi. Diaz is held back by losses to Clay Guida and Joe Stevenson. This is appropriate: while Diaz was impressive in beating Gomi, he’s still just 4-5 in his last 9 fights.

Tony Ferguson stays stuck at 23.23 despite beating Aaron Riley. If Ferguson’s TUF victory over Justin Edwards was counted as a professional fight, his SILVA score would be significantly higher, but Riley isn’t considered a strong enough opponent to boost Ferguson’s score. I firmly believe that SILVA underrates Ferguson, and if Ferguson is matched up against an opponent in the 24-28 SILVA range, I’ll be inclined to pick against SILVA again.

Aaron Riley’s SILVA score is also unchanged at 24.49. Unlike Ferguson, I feel that this properly evaluates Riley. Fighters with a SILVA score that low generally don’t survive in the UFC, and Riley’s UFC record is now just 3-5.

Featherweights

Junior Assuncao’s SILVA score rises only slightly as a result of beating Eddie Yagin, from 20.18 to 21.91. The reason for this is simple: the win over Yagin doesn’t come close to making up for losing to Torrance Taylor earlier in his career. Assuncao will need to beat a more impressive opponent to really rise in the SILVA ratings.

Yagin stays put at 26.38 despite the loss, as SILVA really likes his win over Joe Soto in Bellator, but already penalizes him for losing to Diego Saraiva. The loss to Assuncao is not rated nearly as poorly as the loss to Saraiva, so there’s no reason to reduce Yagin’s rating as far as SILVA is concerned.

Bantamweights

Takeya Mizugaki dispatched of Cole Escovedo in somewhat impressive fashion, but beating Escovedo isn’t enough to budge his SILVA score upwards from 27.56. Mizugaki’s best recent win took place against Rani Yahya; he’ll need to improve that if he wants to move towards the top fighters in the division in SILVA.

By the same token, Escovedo’s loss to Mizugaki isn’t rated as poorly as his loss to Michihiro Omigawa; therefore, Escovedo stays put at 27.66.

Hopefully this helps explain a little bit of how SILVA works. If some of the ratings changes don’t make sense to you, keep in mind that just because one fighter beats an opponent, doesn’t mean that the fighter is better than his opponent. Upsets happen all the time in sports (remember when the Buffalo Bills beat Tom Brady and the New England Patriots 31-0?). If SILVA didn’t change a fighter’s rating, it’s usually because the result wasn’t “surprising” enough.

Advertisements

2 responses to “UFC 135: SILVA Update

  1. howard morton September 25, 2011 at 10:15 am

    I like your work. But SILVA needs a lot of work. Gomi and Hughes losing were so obvious I am kicking myself for not betting more on Diaz and Koscheck. I wanted to bet those fights under 2.5 rounds, but I could not find any casinos taking that action.

    Anxious to hear SILVA on ufc on versus 6 on saturday.

    • David Williams September 26, 2011 at 10:26 pm

      SILVA does indeed need a lot of work, it’s in “version 1.0,” so it’s far from a finished product. And I agree, SILVA should have favored both Diaz and Koscheck, although I don’t think a Koscheck win was “obvious.”

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: