Tuesday, December 20, 2005

Dear Fox Sports:

Dear Fox Sports:

While I find the computer-simulated 16-team playoff amusing, I have to ask: may I have some of whatever drugs the statisticians who programmed the simulation are smoking?

Seriously. Let's take a look at the hypothetical "Penn State-Tulsa" matchup. Austin Scott with 5 carries? The only games in which Austin Scott had more than 5 carries were the Minnesota, Illinois, and Cincinnati games. You know, the ones where the average margin of victory was thirty-seven points. And Scott's carrying in the first quarter? Yah, okay.

So, what went wrong? What's wrong with this simulation? What else is going on? Let's see. Penn State's only rushers are Robinson (151 yards), Hunt (119 yards), and Scott (20 yards). That's weird. In essentially every game this year, Justin King has also rushed for a few. Why would he be missing? Oh - probably because you probably forgot to move him back over to offense after removing Derrick Williams.

That's not all that's strange, though. The only sacks listed are Nolan McCready (who dropped behind Lowry in the offseason this year) and Chris Harrell, who didn't have a sack all year (and wasn't really likely to get one, as we almost never safety blitzed). Not one listed for the defensive line or linebackers - strange.

Plus Robinson passes 10 times, for only 87 yards? That's weird, because he hasn't had that kind of yards-per-completion since the South Florida game. The rest of the season he's been averaging 15 yards per completion. And the long is 16 yards here?

So let's put all this together. Mixing Scott and Hunt. No WR runs. Short yardage passes. What have I just described? Oh, yeah. Last year's team.

From the website, you list WhatIfSports.com as the partnership for this, which is probably where the simulations were run. But if I had to guess, it looks like the simulation was based on a preseason evaluation of Penn State (which... is pretty much what the NCAA games for console systems are). Well, if that's true, we should see it in the results of the games: oh wait, we do! There's Ohio State running roughshod over its competition, Notre Dame also struggling, and Florida State demolishing Oregon. (Penn State, Notre Dame, and Oregon, were the three teams who ended very highly-ranked that were missing at the beginning of the season)

So forgive me if I'm wrong. If it's all just a crazy, wacky coincidence that the box score, along with the players listed, and the tendencies of the teams looks disturbingly like what someone not that knowledgeable about the team would've predicted from 2004. But if you're going to have fun, and do the whole "what if" thing, at least, y'know, have the decency to base it on this season. We know USC, Texas, and Ohio State were touted as the bee's knees at the beginning of the year. Thank you.

But guess what? Some teams didn't live up to their expectations. And others far exceeded them. How bout joining the rest of us in reality?

Thanks so much,
Bleed Blue 'n White

(Seriously, Penn State versus Tulsa? The Tulsa that Minnesota destroyed at the beginning of the year?)

Edit: Fox Sports also has video game simulations of the outcomes of the bowls. Not surprisingly, Florida State (preseason #11 or so) crushes Penn State. Never would've guessed. Tell me again, why don't they come out with video games after the freaking season is over?

2 Comments:

Blogger Mike L said...

Fox Sports = oxymoron

7:24 PM

 
Anonymous Just Passin' Through said...

Why do they base NCAA games on the previous season instead of the current one? Easy: Christmas.

Realistically, you have two possible release dates for all the college football games: right before the season starts, or right after the season ends. If you went with when the season ends, the game would probably come out in mid-February or early March. Now, those of us who are football nuts, we'd love this idea because we could tide ourselves over until the following year with our football game. However, the publisher would take it on the chin for sales because the game would be out on shelves for nine or ten months before the Christmas holiday season, when most games make the real cash. As a result, no games are made then. Another minor point is that if you made your release target in February, you might miss where recruits end up. As a result, your release might get pushed back to mid-April.

Now, by releasing before the season, the game publisher has two advantages: 1. The passion of college football is hitting on all cylinders as everyone gears up for another season (and allowing fans to play out the season along with, or before their team takes the field); and, 2. Christmas is only ~4 months away, and the magazines can hype the game significantly so WalMart, Best Buy, and all the other major retailers can get it in their heads to put the games on the shelves.

The downside of this, as you point out, is that you are left with stats that are based on last year's actions. For professional-based sports, this is okay because you think that most pros won't radically change their game style or ability year-to-year (yes, I realize it does happen when a team changes its offensive/defensive concept, but overall, most pros don't change that much). Unfortunately, for college sports, in particular football, the year-to-year progression of the athletes is significant. As such, basing your stats on how last year's team did is pointless at best and foolhardy at worst (Will that blue-chipper actually pan out? Will he contribute immediately, or will he take time to mature?).

Still, it's what we must live with if we want any games at all.

5:53 PM

 

Post a Comment

<< Home