Geeksheet plays - wk 4 college

thegeeksheet · 09-19-08, 02:19 PM

write ups

03:00PM 377 Arizona Wildcats 3
Saturday 09/20/2008 378 UCLA Bruins

Arizona had a bit of a hiccup losing on the road last week to New Mexico; however, they turned the ball over 5 times and actually outgained NM by a fairly decent margin (particularly considering a -4 TO margin which essentially gives NM 4 extra possessions). Other than that game (where Arizona’s defense allowed 330 ypg +/-) the Zona defense has been very good this year, and even including that game the Arizona defense is very good across the board (+0.7 ypr and +1.5 ypa). I can’t see an abysmal UCLA offense averaging an unbelievable 0.8 ypr this season (29 yards rushing in week 1 and 9 yards rushing last week) with a young inexperienced QB and a new offensive system doing much this week, which is reflected by a 6.1 ypa and 4.1 yppl average this season (263 ypg)? UCLAs offense is outmatched in every facet and the low end of my statistical projections actually call for negative yardage on the ground for UCLA. The statistical mean of my projections call for less than 200 yards offense for UCLA, but to be conservative I doubled the statistical mean for rushing ypr and increased the passing production by 50%, and I still come up with only 235 yards offense for UCLA. We actually have pretty good data on both teams, although only 2 games on UCLA, so I feel pretty good about my mathematical projections. Our models are all pretty consistent with scoring projections for UCLA in the 27-20 range. I think that is a little on the high side, and isn’t totally accounting for how horrible UCLA has been rushing the football, and the matchup challenges they will face this week vs an Arizona defense that has held 2 of its 3 opponents to under 100 yards. I just can’t see the UCLA offense getting on track without a solid ground game. My math calls for 7-10 points (but that is just based on this year). If I adjust 33% for last year (because of only 2 games data) I come up with a 17-22 scoring range which starts to look more like our other models. I’ll hedge on the high side, but think that is the top of expectations for UCLA.

UCLAs defense hasn’t been too bad this year, averaging +0.4 ypr and +1.6 ypa this season, despite getting torched for 51 last week. However, Arizona has been equally good on offense and is actually a statistical draw with UCLA’s Defense (absolute advantage / disadvantage for this side of the ball was 0 for yppl, +0.2 ypr for Arizona, and -0.5 ypa for Arizona). My projections call for 5.8 yppl (which is middle of my high / low projections and actually almost dead average of the two teams average thus far – which isn’t always the case and isn’t a calculation in my projections but in this case with similarly skilled sides is a pretty good validator that we are in the neighborhood). 428 yards offense is almost double the high end of my projections for UCLA, and my models call for 24-30 points for Arizona, and based on just this years numbers I show a high of 52 for Arizona. All in all I can’t see where UCLA stays in this game with out a lot of fortunate ST and TO help. However, only having 2 games data on UCLA will keep me tempered on this one, but only slightly.

Arizona 31 UCLA 20 Arizona is a # 4 recommended play at -3.

07:00PM 341 Utah Utes 9
Saturday 09/20/2008 342 Air Force Falcons

AF continues to be under-valued and I don’t mind backing them again with points in a game at home which looks to be a pretty even draw. AF has been very good running the football (averaging 358 ypg on the ground) at a 5.1 ypr clip thus far. They should get their biggest test of the season vs a Utah team that is only allowing 2.3 ypr, but that is skewed some as they’ve faced some pretty paltry rushing attacks (offensive averages they faced are 111 ypg at 3.2 ypr). AF last year torched Utah for 5.4 ypr on the ground, and while AF isn’t as strong as they were last year on offense, Utah has its own question marks, so I don’t see why we shouldn’t expect similar results. My math calls for 4.8 ypr for AF, which I think given last years performance vs Utah and their performance this year is a pretty accurate assessment. AF has virtually abandoned the pass this season (averaging 6 atts / game), which is probably good here as Utah has been very good vs the pass this year (+1.3 ypa allowing less than 5 ypa), but I don’t see that having a material affect on the AF offensive production (AF was 0-7 last week and still ran for almost 400 yards and 31 pts against a pretty good Houston team). My math calls for 360 total yards (335 on the ground) and 25-28 points. My models have nearly the same spread (23-29 points), and given my math and AFs performance this year and in this series I think that is pretty accurate.

AFs defense actually matches up pretty good vs Utah, as they are below average (-0.4 ypr and 3.8 ypr) defending the rush, but get a Utah offense that has struggled there as well (-0.2 & 3.9 ypr). AFs strength this season has been defending the pass, where they are +0.7 ypa and only allowing 5.4 ypa this season. Utah has been very good in the air averaging 8.1 ypa (+0.7 ypa), so it appears Utah won’t have any real match up angles they can exploit. Good passing teams on offense generally perform better vs good passing defenses (when compared to the rushing side which is reverse), so I still think Utah should have some success throwing (my math calls for 6.8 ypa for Utah). AF allowed 6.4 ypa last week to a similar offensive team in Houston, so that seems pretty fair. All in all I see about 360 yards offense for Utah and about 21-28 points. My models are pretty consistent with Utah at 22-25 points.

I’ll take AF at home with nearly double digits in a game in which my multi-situational simulator shows them winning outright about 30% of the time, and in which they should have the advantage on both sides of the rushing attack. AF has been very good to us thus far, and continues to be under-rated. Given all the above I don’t mind backing them one more time. Every one of our models show AF winning outright, and our worse case scenario shows a 25-21 Utah win.

AF 24 Utah 25 – AF is a #4 recommended play at +9

07:30PM 313 Miami Ohio Redhawks
Saturday 09/20/2008 314 Cincinnati Bearcats 11.5

This one is a little tricky as we really only have 2 games of data for Mi(oh) (ignoring Charleston Southern last week) and really ½ a game on Cincy (they’ve played Eastern Kentucky which I will largely ignore and at Oklahoma which is one of the top 3 offenses in the nation). Still all those things considering, I think Cincy is the side here. Mi(oh) has really struggled vs Vandy and Michigan to move the football, and expect them to do so here where they are at nearly a 1.0 ypr and yppl disadvantage. My projections call for only 4.8 yppl (including 2.7 ypr). The tricky part about this game is Oklahoma lit up Cincy’s defense last week in the air to the tune of 9.7 ypa. Last year Cincy allowed 6.8 ypa including just 4.9 ypa to Mi(oh), and they return 4 of their back 7, so while their should be some drop off from last year, it shouldn’t be much. When I adjust for this (vs using just this years Cincy defensive stats which is just Oklahoma) I get a projection of 5.8 ypa for Mi(oh) and 4.2 yppl which seems more realistic to me. All in all my projections call for about 260-310 yards offense for Miami and about 10-17 pts, which given the fact they allowed 10 last year to Miami and averaged allowing 18ppg last year seems right.

Miami’s defense cannot stop the run and is allowed -0.7 ypr to Michigan and Vandy which actually rate out about the same to a little worse than Cincinnati (my preseason projections call for Cinc to average 3.8 – 4.0 ypr compared to 3.5-3.8 range for VU and Mich). Even last week vs Oklahoma they averaged a respectable 3 ypr. My guess is they’ll find the landscape much easier to navigate this week vs a pretty weak Miami defensive front. My projections call for a +1.4 ypr advantage for Cinci and 4.3 ypr. Miami hasn’t been very good vs the pass either and Cincy has nearly a +1.0 ypa advantage in that dept as well. All in all my projections call for 406 yds offense (which considering Cinc averaged 430+ last year and had 438 last year vs Miami seems fair) and 28-33 pts. All 5 of our models project 35-40 points here which is very consistent with my projections and models. Cincy seems to own every advantage in this game, and think they cruise to a pretty comfortable victory. Even if they only perform as well as they did vs Oklahoma (which presumes Miami’s defense is as good as Oklahoma which they clearly are not) I think there is a decent chance you squeak out a cover as I have a low projection for Miami of 245 yards offense and 10 pts. I think the more likely scenario is about 285 and 16 for Miami and 420 and 36for Cincinnati.

Cincinnati 36 Miami 16 – Cincinnati is a #4 recommended play at -11.5.

12:00PM 333 Iowa Hawkeyes 1
Saturday 09/20/2008 334 Pittsburgh Panthers

This is a very interesting game as we don’t really know a whole lot about either teams offense, and very little about Iowa in general after two tune ups vs 1AA Maine and FIU followed by a very poor offensive performance vs an average but by no means spectacular Iowa State defense. Still Pitt hasn’t shown its defense is going to be a rock this season without DC Paul Rhodes (now at Auburn), allowing 22 ppg to BG and Buffalo (not necessarily offensive juggernauts).

There is a lot of extrapolation that goes into this game, which increases the margin for error; although in basically a pick em game, with a very proven and stable Iowa Defense, I don’t mind backing Iowa here as I think all the upside is with them. Pitt can’t throw the ball, and has a -1.0 ypa disadvantage vs a good Iowa pass unit (allowing only 5.0 ypa this season), and with no real rushing game to speak of (Pitt is only averaging 3.6 ypr and get a defense very good vs the rush and has a -0.6 ypr disadvantage) I am not sure what they are going to be able to do to move the football. My math calls for 284 yds offense (3.0 ypr & 5.0 ypa) and a scoring range of negative 0.5 pts to 11 pts. Iowa’s defense is far superior to BG and Buffalo so even if you only discounted Pitts performance vs those units by 15-25% you still could only presume 14-18 pts for Pitt. I’ll conservatively give them that, although I’ll be surprised to see them get double digits. My models are all in that range as well (right at 18-19).

Iowa really struggled last week in new QB Ricky Stanzi’s 1st real test. I actually like that they were challenged last week as it gives the coaching staff an entire week to digest what he was successful at running, and what gave him troubles when under a little pressure. This will be his 1st road test, so I expect a slow start and some bumps along the way. Pitts defense has been very good this year (allowing 4.4 yppl), but again vs BG and Buffalo it’s difficult to know what they really have. My math calls for a little over 5.1 yppl and 3.7 ypr on the ground (7.0 ypa) for about 375 yards offense. All my models have Iowa at 20-24 and my math calls for 20-27. Even if Pitt’s defense comes to life and performs better than it did vs weaker offenses (they allowed 22ppg to BG and Buffalo) and by a drastic margin (again say 25%), you still come up with a worse case scenario of 16-18 pts which is the best case scenario for Pitt. With all the model alignment, math, statistics, and match ups supporting Iowa I’ll forgo the lesser than desirable line value and make Iowa a recommended play.

Iowa 23 Pitt 14. Iowa is a #4 recommended play at -1.

Free Selections

Miss St @ Georgia Tech (-7)
Miss State’s only shot to score any points is to get something going on the ground. This year vs the 2 D1 schools they’ve faced they have yet to eclipse the 100 yard mark. GTs defense is slightly better than average vs the run (+0.4) and my math projects 2.8 – 3.1 ypr for Miss St (barely 100 yards on the ground). State is horrible in the air and Tech has been fantastic defending the pass (allowing 3.7 ypa on defense) and my math calls for 228 yards total offense for Miss St.

My problem with this game is that Miss St has been (and under Croom have historically been very good) defending the rush. This season they are only allowing 2.8 ypr, and while GT has put up some big numbers on the ground (averaging 263 ypg and 5.8 ypr) they are doing vs pretty bad defenses that are allowing 5.4 ypr. GT has actually been very good passing, and PJ has shown a willingness to adapt his 95% run mentality to his personnel (he’s running a 80-20 run-pass ratio thus far). I wish he would open it up a little more here as he has nearly a 2.0 ypa advantage when they pass vs Miss St. Still my projections only call for 12-15 atts which means about 310-335 yards total offense for GT. Laying 7 pts is just too much for me here in a game the average margin of victory for GT across all my models and math is 11 pts. (25-14). Given what I’ve seen from Miss St (and my math projections) I think 14 pts and 2-3 scores is way to generous for State, but there is too much risk with the GT rushing situation, so I’ll pass. GT 21 Miss St 10.

ECU (-7) @ NCSU

On paper ECU wipes the floor with NCSU. My problem is I think the ECU paper is being inflated by some wins over very over-rated teams in VT and West Virginia, and although NCSU is allowing 28 ppg they are only giving up 367 ypg and 5.6 yppl which is actually pretty good (ECU is only averaging 357 and 5.4). I think ECU is getting way over-valued here, and last weeks struggle vs a pretty bad Tulane team is starting to show it. There is no way I can back an NCSU team that hasn’t show any ability to move the football, but I think that 7 is too much to lay for an ECU team making their 1st real road start in a hostile environment (they even struggled last week at Tulane which is far from a real road game) vs an NCSU getting their 1st real meaningful home game with a chip on their shoulder and something to prove. All my math has ECU winning 29/32- 18/20 but something gives me the feeling this thing could be close, and I want no part of it. Still I am a math guy, and while I won’t let situations determine a play for me, I will certainly let it take me off a play. ECU 27 NCSU 17

UGA -6.5 @ Ariz St

The Sundevils got caught last week looking ahead to this game, and UGA was exposed slightly in their 1st real road test of the season @ South Carolina vs a very good USC defense. Statistically it is a very similar game 7.1 yppl (+0.5 yppl) offense vs 4.1 yppl defense (+0.1 yppl) for the UGA O– ASU D and 7.0 yppl offense (+0.6 yppl) vs 4.6 yppl (0.2 yppl) defense for the ASU O vs UGA D. This early on when you have this wide a variance in the sides it leads to wide margins in projected totals which we got for both sides (7.0-4.7 yppl for UGA and 5.2-6.8 yppl for ASU).

All my models have 28-30 pts for UGA and 20-24 for ASU, given the above there is just a little too much uncertainty for me to back either team here even though all my models have UGA covering by 5-9 pts (although because of some of the variances you get a worse case scenario where ASU outgains UGA 407-344). I’ll pass but I do think UGA will get the job done. UGA 30 ASU 20.

Rutgers (-4.5) @ Navy

My math and models have been really close to Navy every week this season, but their uncertainty on defense and at the QB spot has kept me off them every week and that is again the case this week. I have no idea what is going on with Rutgers, and that adds even further to the uncertainty with this game this week. There is no need to repeat what I’ve said about Navy the last couple weeks as the same thing is at play here. Navy probably gets 300+ on the ground but will get torched in the air (this week I look for 496 yds Navy – 403 Rutgers), Navy probably gets to 30 points, but the way their defense is playing could give up 40 (although Rutgers has looked horrible thus far averaging 9.5 ppg). One has to think Rutgers wakes up from this funk, and Navy’s defense is surely medicine for that. Still all my math, models, and analysis show that Navy should win this game, but given the intangibles and Navy’s horrific pass D I’ll wait another week for a better spot. Navy 34 Rutgers 30

M.W. · 09-19-08, 03:09 PM

Good luck.

thegeeksheet · 09-22-08, 08:58 PM

3-1 this week (7-2 last 2 weeks since i started posting). prety good start to the season, no real complaints on Iowa game, had the projected yardages and yppl damn near perfect (See write up below and go ck the box scores).

will check back in later this week with thoughts on this weekend. i sort of peg this weekend and next as the start to my season based on my math models b/c i have atleast 3-4 quality games of data for 90% of the CFB teams this season.

good luck to all.

Originally posted by thegeeksheet

All #4 rec plays
Ariz -3
AF +9
Cinc -11.5
Iowa -1

4-1 last week (1st week i posted)...here is link

http://forum.sbrforum.com/college-football-handicapping/100681-my-week-3-plays.html#post1031427

will post write up in seperate with others i came close to but passed on...good luck to all