Predictor Week 8

LegendaryGT

Dodd-Like
Joined
Oct 18, 2009
Messages
62,557
Predictor said:
GT vs FSU
Week 8 of 2015

GT - Offense
Yardage: 379 ypg
Performance: 15 (higher is better)
Scoring: 31
Efficiency: 11 (lower is better)
Performance: 22 (higher is better)
GT - Defense
Yardage: 379 ypg allowed
Performance: 2 (higher is better)
Scoring: 31
Efficiency: 12 (higher is better)
Performance: -5 (higher is better)

FSU - Offense
Yardage: 410 ypg
Performance: 11 (higher is better)
Scoring: 28
Efficiency: 14 (lower is better)
Performance: 9 (higher is better)
FSU - Defense
Yardage: 333 ypg allowed
Performance: 6 (higher is better)
Scoring: 15
Efficiency: 22 (higher is better)
Performance: 31 (higher is better)

Prediction:

Pure statistical: GT 22 (378 yards), FSU 31 (411 yards)
GT best case: GT 31 (413 yards), FSU 23 (389 yards)
FSU best case: GT 15 (353 yards), FSU 37 (438 yards)

GT talent: 5051
FSU talent: 12366
GT talent of schedule: 6951
FSU talent of schedule: 5881

Suggested talent adjustment: 62% towards FSU best case
Final: GT 18 (362 yards), FSU 35 (428 yards)

Seems reasonable.

GT 48, UVA 25
GT 31, VT 31
GT 33, MIA 40
GT 34, UGA 42

3-9 or 4-8 depending upon how OT goes with VT. It favors VT by 9 yards in regulation.

Predictor Poll said:
#1: ('OSU', [0.988, 122, 3, 7, 0])
#2: ('CLEM', [0.984, 121, 4, 5, 0])
#3: ('LSU', [0.976, 119, 6, 6, 0])
#4: ('FSU', [0.968, 117, 8, 6, 0])
#5: ('BAY', [0.964, 116, 9, 5, 0])
#6: ('UTAH', [0.948, 112, 13, 6, 0])
#7: ('IOWA', [0.948, 112, 13, 6, 0])
#8: ('ALA', [0.9286, 125, 0, 6, 1])
#9: ('FLA', [0.9126, 121, 4, 6, 1])
#10: ('OKLA', [0.8927, 119, 6, 5, 1])
#11: ('TCU', [0.892, 98, 27, 6, 0])
#12: ('ND', [0.8846, 114, 11, 6, 1])
#13: ('MSU', [0.88, 95, 30, 7, 0])
#14: ('HOU', [0.868, 92, 33, 5, 0])
#15: ('TOL', [0.86, 90, 35, 6, 0])
#16: ('MICH', [0.8531, 124, 1, 5, 2])
#17: ('TA&M', [0.8527, 109, 16, 5, 1])
#18: ('STAN', [0.8527, 109, 16, 5, 1])
#19: ('TEM', [0.816, 79, 46, 5, 0])
#20: ('MEM', [0.816, 79, 46, 5, 0])
#21: ('WIS', [0.8091, 113, 12, 5, 2])
#22: ('UNC', [0.807, 108, 17, 3, 1])
#23: ('CAL', [0.804, 101, 24, 4, 1])
#24: ('OKST', [0.776, 69, 56, 4, 0])
#25: ('MSST', [0.7613, 107, 18, 4, 2])

#76: ('GT', [0.3993, 79, 46, 1, 5])

It still ranks us higher than #79 SYR, #81 BC, #90 WAKE, and #93 UVA, in conference. The ACC be so deep.

I haven't looked at the schedule for next week. Any requests?
 
GT best case: GT 31 (413 yards), FSU 23 (389 yards)

giphy-facebook_s.jpg
 
Meanwhile, Sagarin predictor still picks us to beat every team left on our schedule except FSU, with UGA as a virtual toss-up.
 
I just had a thought...

Are we the new Arkansas? Overrated as hell, loses almost every game in conference, but computers love us and we get credit for losing to good teams.
 
Predictor is off on VT I think. Does it account for the return of Brewer?

Nope, and it also isn't using week 1 as a data point any more, which takes VT's game against Ohio State out of the stats. Next week we'll lose Tulane as a data point and it will start to look really grim for us, I suspect.
 
I just had a thought...

Are we the new Arkansas? Overrated as hell, loses almost every game in conference, but computers love us and we get credit for losing to good teams.

I already called this last week.
 
Nope, and it also isn't using week 1 as a data point any more, which takes VT's game against Ohio State out of the stats. Next week we'll lose Tulane as a data point and it will start to look really grim for us, I suspect.
It maybe too much work at this point, but you might want to try tapering off older games instead of completely throwing them out.
 
Nope, and it also isn't using week 1 as a data point any more, which takes VT's game against Ohio State out of the stats. Next week we'll lose Tulane as a data point and it will start to look really grim for us, I suspect.

Yeah, but it'll take into account our massive beat-down of FSU so those should about even out, no?
 
I just had a thought...

Are we the new Arkansas? Overrated as hell, loses almost every game in conference, but computers love us and we get credit for losing to good teams.

In the full-data round robin it still picks us to beat more than half the FBS. In the round robin with data from the prediction window (which is everything but week 1 right now), it actually likes us just slightly more (81 wins vs 79).

Excluding week 1 data, here's what the poll looks like:

Predictor Poll said:
#1: ('OSU', [0.988, 122, 3, 6, 0])
#2: ('CLEM', [0.988, 122, 3, 5, 0])
#3: ('BAY', [0.972, 118, 7, 4, 0])
#4: ('STAN', [0.968, 117, 8, 5, 0])
#5: ('LSU', [0.96, 115, 10, 6, 0])
#6: ('FSU', [0.96, 115, 10, 5, 0])
#7: ('IOWA', [0.952, 113, 12, 6, 0])
#8: ('UNC', [0.948, 112, 13, 3, 0])

#9: ('UTAH', [0.94, 110, 15, 5, 0])
#10: ('ALA', [0.9167, 125, 0, 5, 1])
#11: ('MICH', [0.9127, 124, 1, 5, 1])
#12: ('TCU', [0.908, 102, 23, 5, 0])
#13: ('MSU', [0.9, 100, 25, 6, 0])
#14: ('FLA', [0.8967, 120, 5, 5, 1])
#15: ('WIS', [0.8847, 117, 8, 5, 1])
#16: ('HOU', [0.884, 96, 29, 5, 0])
#17: ('PSU', [0.8687, 113, 12, 5, 1])
#18: ('TOL', [0.868, 92, 33, 6, 0])
#19: ('OKLA', [0.86, 115, 10, 4, 1])
#20: ('ND', [0.8487, 108, 17, 5, 1])

#21: ('MEM', [0.84, 85, 40, 5, 0])
#22: ('TA&M', [0.82, 105, 20, 4, 1])
#23: ('CAL', [0.82, 105, 20, 4, 1])
#24: ('GASO', [0.78, 70, 55, 4, 0])
#25: ('PITT', [0.76, 90, 35, 4, 1])

#32: ('DUKE', [0.735, 90, 35, 3, 1])

It wants to think that part of the reason we look so bad is because we have had a brutal schedule.
 
It maybe too much work at this point, but you might want to try tapering off older games instead of completely throwing them out.

I think it wouldn't actually be too much work, but I have no idea if it would improve the accuracy.
 
Back
Top