Nearly a year ago, I proposed an idea to the community here at ODB about doing projections for the major offensive and defensive players on Baylor's 2013 football team. The basic concept was that members here would project the notable statistics for each player, I'd compile and compare them, and then someone would win sweet, sweet swag. Last week, while trying to lay out the plan for this year's season preparation, I realized once again that I hadn't ever gotten around to the compiling and comparing part. So I took way too many hours to do it this afternoon, evening, and night.
How This Works:
Before I get into the meat of what you came here to see, I'm going to pull a Morgan Freeman from Batman Begins for a second: this was a ton of work. Those of you who have been around here for a while know that I love spreadsheets. I have dozens of them for ODB alone. This was by far the most time-intensive one I've ever done for the blog, and by far the least fun. I'll just be honest. I had to pull the numbers from each thread and input them into the spreadsheet manually, then go back to my MIS classes from a decade ago to remember how to do what I wanted to do in Excel. Eventually, I settled on z-scores around the actual results with each player weighted evenly. In that situation, you want as low a score as you can get, 0 being the lowest possible.
After I had each person's Z-Score for each player, I put them into the "Rankings" page (second in the book below) and averaged out each person's score. Whoever had the lowest average Z-Score across all 11 players won. It ended up being a very tight race, as you already know if you were on Twitter this evening for #livetweetthespreadsheets
A few observations:
- 35 ODB members made at least one projection for a player last season, including 9 that made just one. Three of those just did the very first one for Petty and then sat out the rest. I included in the spreadsheets everyone that did even one projection, but I had to limit eligibility for being named the overall winner (for offense) to just those people that did all 11 offensive players. Otherwise, you get into situations where someone could have one really low score and win where they shouldn't. Of the 35 who submitted projections, 17 did all 11 players. That's actually higher than I expected given that this is the first time we've ever done this.
- In putting this together, I did each player separately before going back and doing the "Rankings" page in the linked spreadsheet. The whole time, I thought I had a pretty good handle from looking at the numbers who was going to win. I didn't. I was convinced the entire time that deafkittens, dfw mike, or sic'em had this thing in the bag and that the person who eventually won wasn't going to be close. Welp.
- It was kinda funny to see when people typically responded during the course of the day. Had I more time, I might work out the patterns that I'm pretty sure exist there. I know I saw several situations where two posters did projections right before or after each other over the course of days.
- dfw mike actually projected twice on Lee, Rhodes, Norwood, and Fuller. Each time, I flipped a coin to figure out which one to drop. At first, I was a little annoyed because my numbers didn't match up, then it started being funnier the longer I stared at the computer.
- We as a community really missed on a couple of guys, as you'll see, but also absolutely nailed a few. I'll get to that more below, but I was impressed.
- There are a couple of situations where the data probably doesn't fit what you expected, namely with Petty and Seastrunk, and I'll explain why below.
The Top 5 Non-Fuller* Individual Projections:
Ted Harrison -- Coleman -- .4645
sic'em -- Norwood -- .4991
HUG ME RON -- Martin -- .5067
sic'em -- Najvar -- .7261
J.A. Grace -- Petty -- .7497
*You'll see why below.
QB Bryce Petty | Best Projection: J.A. Grace .7497 | Community: .9119
Right off the bat, we have one of those weird situations I talked about earlier. While ODB did what I would call a fantastic job on the whole projecting Bryce's passing stats, we hilariously overrated his rushing yards this past season. We were so off on the whole that I had to take rushing stats out of it completely; otherwise whoever put the lowest numbers there was almost assured to win. I realize that's different than I originally said, but I think it's more fair this way so one stat doesn't tank basically the whole thing. And it probably would have.
Top 5 Projections for Bryce:
Completions | Attempts | Yards | TDs | INTs | |
ACTUAL RESULTS | 250 | 403 | 4200 | 32 | 3 |
J.A. Grace | 240 | 405 | 3850 | 33 | 8 |
deafkittens | 252 | 370 | 3800 | 32 | 7 |
sic'em | 267 | 400 | 4027 | 28 | 6 |
Baylor Genin37 | 281 | 401 | 3950 | 36 | 5 |
Aqua/Seymour | 261 | 380 | 3884 | 32 | 10 |
As I said, I was really happy with ODB's average projection for Bryce's passing stats.
RB Lache Seastrunk | Best Projection: Quintas11 -- 1.409 | Community -- 2.913
This was the other situation where I had to change things from the beginning. As a community, ODB was extremely bullish last season on LacHeisman's numbers across the board, but even more so in his potential as a pass-catcher. The average ODB user projected 23 receptions for 244 yards and 4 TDs for Lache with some going much higher. He obviously caught zero passes for zero yards and zero TDs. That created a situation where the winner was once again whoever put the lowest number in his pass stats, and since that turned out to be someone who just didn't include them at all, I took them out completely.
Top 5 Projections for Lache:
ATT | YARDS | TDs | |
ACTUAL RESULTS | 158 | 1177 | 11 |
Quintas11 | 211 | 1212 | 14 |
TNBear | 210 | 1460 | 12 |
RSWaco | 208 | 1435 | 8 |
deafkittens | 210 | 1600 | 14 |
mileskelly | 225 | 1645 | 13 |
Injuries essentially determined this one.
RB Glasco Martin | Best Projection: HUG ME RON -- .5067 | Community -- 1.4706
The first so far where someone really knocked it out of the park. HUG ME RON made a strong statement in projecting Glasco lower than many others, some of whom actually had him over Lache in their projections. Injuries sadly kept Glasco out much of this past season, limiting his production and allowing others to step up in his absence. I kept his receiving numbers in for two reasons: 1) it's hilarious to me he caught a pass and Lache didn't (and what a pass it was!), and 2) the numbers were so small it didn't make much of a difference.
Top 5 Projections for Glasco:
ATT | YARDS | TDs | CATCHES | YARDS | TDS | |
ACTUAL RESULTS | 120 | 509 | 7 | 1 | 25 | 0 |
HUG ME RON | 120 | 680 | 9 | 2 | 8 | 0 |
deafkittens | 140 | 750 | 12 | 2 | 5 | 0 |
sic'em | 142 | 697 | 12 | 0 | 0 | 0 |
RSWaco | 110 | 800 | 13 | 0 | 0 | 0 |
TheEuphoriac | 160 | 800 | 12 | 2 | 8 | 0 |
TE Jordan Najvar | Best Projection: sic'em -- .7261 | Community -- 1.853
Nothing much to say here. sic'em and bgoodrich nailed this one, but hardly anyone went insane projecting a huge year for Najvar. Since he didn't have one, that was good for the average.
Top 5 Projections for Najvar:
CATCHES | YARDS | TDs | |
ACTUAL RESULTS | 10 | 85 | 0 |
sic'em | 12 | 90 | 2 |
bgoodrich | 12 | 103 | 2 |
corey.royse | 13 | 100 | 2 |
Baylor Genin37 | 11 | 98 | 3 |
mileskelly | 14 | 140 | 2 |
Those who didn't project too many TDs were probably in good shape.
WR Tevin Reese | Best Projection: _007 -- 3.0327 | Community -- 4.0078
There's good news and bad news. The bad: ODB completely whiffed projecting Reese, as evidenced by the fact that a projection averaging 3 standard deviations from the actual result was the best we had. The good news? It was almost definitely a result of injuries keeping him out the last few games. Since injuries are almost impossible to predict, it doesn't bother me.
Top 5 Projections for Tevin:
CATCHES | YARDS | TDs | |
ACTUAL RESULTS | 38 | 867 | 8 |
_007 | 65 | 1150 | 12 |
hoop_er | 68 | 1244 | 10 |
sic'em | 65 | 1180 | 12 |
RSWaco | 66 | 1250 | 11 |
theEuphoriac | 70 | 1150 | 13 |
See what I mean? Even the best projections nearly doubled his actual number of passes caught. That's a function of his injury.
WR Jay Lee | Best Projection: sic'em -- .4991 | Community -- 2.2538
ODB was hilariously bad on two players from this past season. Lee was the first. It ended up being that if you went low, you scored really well. If you didn't, you probably took yourself out of the entire thing.
Top 5 Projections for Lee:
CATCHES | YARDS | TDs | |
ACTUAL RESULTS | 22 | 309 | 1 |
sic'em | 32 | 317 | 2 |
dfwmike | 29 | 480 | 5 |
TNBear | 40 | 450 | 4 |
Colby Conner | 47 | 423 | 4 |
RSWaco | 40 | 550 | 5 |
This was about the time when I started thinking sic'em was going to run away with this thing. He'd won two players out of the first six and was firmly in the lead as I gathered the stats in the Rankings page.
WR Robbie Rhodes | Best Projection: dfw mike -- 1.2793 | Community -- 3.0498
If Lee was one guy ODB bit hard on without success, Rhodes was the other. A ton of people that were doing really well rankings-wise fell off here, since Rhodes only caught 10 passes the entire season. Some people, myself included, were predicting a Kendall Wright-esque freshman season, and that did not come to pass at all. Moral of the story: don't bet big on freshmen WRs unless you have a really, really good reason. Recruiting rankings don't count. If you went high on Rhodes, you were basically out at this point, especially if you went high on Lee before that.
Top 5 Projections for Rhodes:
CATCHES | YARDS | TDs | |
ACTUAL RESULTS | 10 | 157 | 0 |
dfw mike | 26 | 320 | 2 |
imahammeru | 30 | 450 | 2 |
RSWaco | 32 | 450 | 2 |
Colby Conner | 32 | 400 | 3 |
mileskelly | 30 | 390 | 4 |
He was the chosen one...
WR Corey Coleman | Best Projection: Ted Harrison -- .4645 | Community -- 1.7336
Ted led the community back on track in a big way with the projections for Coleman's sophomore year. This was one where people had a pretty strong grip on what Coleman would be in the offense last year, even if wasn't close to our best yet, as you'll see.
Top 5 Projections for Corey:
CATCHES | YARDS | TDs | |
ACTUAL RESULTS | 35 | 527 | 2 |
Ted Harrison | 36 | 428 | 3 |
Colby Conner | 35 | 385 | 4 |
Miles Kelly | 42 | 520 | 5 |
imahammeru | 40 | 600 | 5 |
BaylorM&A | 45 | 627 | 5 |
Attaboy, Ted! One of the names on the list here has now positioned himself firmly in the lead.
WR Levi Norwood | Best Projection: lancebu -- 1.0904 | Community -- 1.7354
What's the deal, lancebu? You only did two projections, but they were both great! Did you not know about the whole thing? If this were Starship Troopers, I'd definitely want to know more.
Top 5 Projections for Levi:
CATCHES | YARDS | TDs | |
ACTUAL RESULTS | 47 | 733 | 8 |
lancebu | 54 | 621 | 6 |
_007 | 45 | 600 | 5 |
theEuphoriac | 50 | 600 | 5 |
deafkittens | 60 | 720 | 5 |
Baylor Genin37 | 55 | 587 | 5 |
This one was actually really close; ODB as a whole did a good job with Levi and the top projections reflect how bunched up things were. Nobody projected him to get 8 TDs on the season, marking the first time the community actually underestimated someone in terms of endzone trips.
WR Clay Fuller | Best Projection: mileskelly -- .2230 | Community -- .3670
If you haven't opened up the spreadsheet link above for whatever reason, do it now. ODB as a whole absolutely nailed Clay Fuller's projections last season, and miles had the best of the bunch. This was the community's Mona Lisa, and it almost makes up for missing so badly on Lee and Rhodes. Almost.
Top 5 Projections for Clay:
CATCHES | YARDS | TDs | |
ACTUAL RESULTS | 32 | 512 | 3 |
mileskelly | 35 | 450 | 3 |
stefan.fittingiii | 40 | 497 | 3 |
_007 | 35 | 525 | 4 |
Baylor Genin37 | 42 | 467 | 3 |
Ted Harrison | 40 | 426 | 3 |
A community z-score of .3670 is incredible.
WR Antwan Goodley | Best Projection: imahammeru -- 2.1025 | Community -- 3.1880
All I can say here is that if you participated in last year's projections and claim to have seen Goodley's breakout year coming, you're a dirty liar. Nobody went higher than 675 yards on a guy that would end up with more than double that last season. I'll give us a pass here since nobody in the media saw it coming, either. But Briles did. Because Briles always does.
Top 5 Projections for 'Twan:
ATT | YARDS | TDs | CATCHES | YARDS | TDS | |
ACTUAL RESULTS | 6 | 33 | 0 | 77 | 1372 | 13 |
imahammeru | 6 | 50 | 0 | 45 | 675 | 8 |
J.A.Grace | 10 | 50 | 1 | 50 | 500 | 10 |
Baylor Genin37 | 10 | 140 | 1 | 51 | 611 | 7 |
Colby Conner | 5 | 35 | 0 | 42 | 420 | 5 |
pbpope | 6 | 47 | 1 | 39 | 475 | 5 |
It probably would have made sense here to separate out rushing yards, but everyone was so far off on receiving that didn't matter, anyway. And I was really tired by this point.
Now, without further ado:
Overall Best Offensive Projector: Colby Conner
Take a look at the spreadsheet for the full results from the offensive side of the projections. Colby, our offensive winner, took advantage of the massive community overshoot on Lee to break into the Top 5, then followed it up with good scores on Rhodes and Levi before solidifying his win with another individual top 5 for Goodley. As I said, when I was doing each individual player, I didn't see this coming at all. It was essentially a three-horse race the entire way with deafkittens, sic'em, and dfw mike pacing the pack, but then they fell off and Colby took advantage. Congrats to Colby for being our best offensive projector last season! We'll work out your swag winnings in coming days/weeks.
The Final Top 10:
Member: | Overall Z-Score | Rank |
Colby Conner | 1.869440339 | 1 |
dfw mike | 1.962809447 | 2 |
deafkittens | 1.999054567 | 3 |
imahammeru | 1.999949178 | 4 |
sic'em | 2.079341541 | 5 |
_007 | 2.13837281 | 6 |
TNBear | 2.151037794 | 7 |
corey.royse | 2.239657967 | 8 |
Baylor Genin37 | 2.283273425 | 9 |
Aqua/Seymour | 2.315974206 | 10 |
I have no idea if/when I'll do the defensive players. Look at the spreadsheets (there's a ton of information I didn't mention in the post) and vote in the poll below so we can decide if we want to do it again!