Iowa State vs Oklahoma State analytics

PSYclone22

Visual Analytics Mercenary
SuperFanatic
Aug 15, 2012
5,084
3,201
113
Des Moines
I've been messing around with some analytics using play-by-play data and thought I'd dump a few things here for people to parse through.

First: 1st down tendencies
In the first half of games OK State is calling a run on 1st and 10 nearly two-thirds of the time. Iowa State is at 57% run ratio in the first half. Not much changes for them in the second half, but we get a little bit more run heavy.

From a yards per play perspective, OK State has been pretty inefficient on 1st downs. That could be a key advantage for us on Saturday.

1603469239342.png

Quarterback:

Spencer Sanders should be back Saturday so I'm not showing anything for OK State, but here's a bit for Purdy.

On 1st downs Purdy is more efficient. And by quite a lot. If we can have a GOOD BALANCE of passing on 1st downs to stay ahead of the chains to ensure a lot of short yardage 2nd and 3rd downs or even 1st down conversions we will be successful. I know the 62 pass attempts from last year has people balking at passing, but look at the above chart: we gain >3 yards more on 1st downs by passing than rushing.

1603469778110.png

RB Comparison:

From a strict yards per carry perspective, Hall outgains Hubbard 6.0 to 4.9.

That said, they both tend to cluster their carries at around 3 yards per carry (Breece's most common yards gained are 1 yard and 2 yards, Hubbard's most common carry gets him 3 yards). The difference though is that Hall has produced a few more explosive plays for long yardage (his 3 longest carries are longer than any carry Hubbard has produced).

1603470396805.png

Receivers

The first couple tables show who each team is targeting on 1st downs. Tylan Wallace garners 50% (!) of Oklahoma State's targets on 1st downs. And there's a good reason. Everyone else has been garbage on 1st downs as receivers for them.

For Iowa State, Xavier Hutchinson leads the way with 8 targets on 1st downs, although I'd hazard to guess Kolar will overtake him here eventually. And whereas the Cowboys only get production from Wallace on 1st down passes, Purdy can spread the ball out to just about anyone and be successful.

1603471428035.png

The final charts here reflect receivers on all plays for each team, measuring completion percentage (color and label), yards per target (length of bar), and number of targets (width of bar).

Landen Akers leads Iowa State with nearly 11 yards per target. The completion percentage to him is lower, but completion percentage is less of a big deal for me than yards per target, since completion percentage can be very depth-of-target dependent.

For Oklahoma State... well, again there's no surprise here. Just cover Wallace, right?

1603471464195.png

Share your thoughts, let me know what you like / don't and even any other visuals or analytics you might be interested in. All of this is produced via play-by-play data.
 

CloniesForLife

Well-Known Member
SuperFanatic
SuperFanatic T2
Apr 22, 2015
15,469
20,817
113
When is Campbell (or CW) going to hire you? Also, where are you getting your data? Nice work!
 
  • Like
Reactions: khardbored

DSMCy

Well-Known Member
SuperFanatic
SuperFanatic T2
Feb 1, 2013
5,689
7,352
113
West Des Moines
Yah this is good stuff. I jumped to the receivers section and have a couple questions.
Why did you use a min of 5 targets for ISU but only 3 for OSU?
Am I reading it correctly that Hubbard is averaging -2 yards per catch? Did he just have 1 catch that went for big negative yards?

This definitely shows ISU/Purdy is really spreading the ball around. Should make it harder for teams to take 1 guy away.

I know it would really skew in ISU's favor, but I'd be interested to see ISU's receiving stats excluding the ULL game. I think we can all agree that was an anomaly at this point.
 
  • Agree
Reactions: khardbored

PSYclone22

Visual Analytics Mercenary
SuperFanatic
Aug 15, 2012
5,084
3,201
113
Des Moines
Yah this is good stuff. I jumped to the receivers section and have a couple questions.
Why did you use a min of 5 targets for ISU but only 3 for OSU?
Am I reading it correctly that Hubbard is averaging -2 yards per catch? Did he just have 1 catch that went for big negative yards?

This definitely shows ISU/Purdy is really spreading the ball around. Should make it harder for teams to take 1 guy away.

I know it would really skew in ISU's favor, but I'd be interested to see ISU's receiving stats excluding the ULL game. I think we can all agree that was an anomaly at this point.
Number of targets were just because of games played. Maybe I can do something like "at least 5% of target share".

I'll check on the Hubbard receiving, but yes it's negative.
 
  • Like
Reactions: DSMCy

JHUNSY

Well-Known Member
Aug 31, 2013
5,294
3,097
113
Des Moines, IA
Thanks, really enjoyed reading.

If you could show zone charts of percentage tendencies (offensive and defensive) as to where they like to go when they run on first down or where they tend to throw to on third down, etc. I think that would be really interesting to digest. Kind of like a shot chart in basketball.

Doesn't need to have near as many zones as this, but something to separate flats, mid-range, and deep balls across left, middle, and right and temperate colors to show their percentage of successful passes in that particular zone, etc.

Zones.gif
 

BryceC

Well-Known Member
SuperFanatic
SuperFanatic T2
Mar 23, 2006
26,405
19,538
113
Honestly something I never considered was yards per target for receivers. That is a great way to look at things.
 

PSYclone22

Visual Analytics Mercenary
SuperFanatic
Aug 15, 2012
5,084
3,201
113
Des Moines
Thanks, really enjoyed reading.

If you could show zone charts of percentage tendencies (offensive and defensive) as to where they like to go when they run on first down or where they tend to throw to on third down, etc. I think that would be really interesting to digest. Kind of like a shot chart in basketball.

Doesn't need to have near as many zones as this, but something to separate flats, mid-range, and deep balls across left, middle, and right and temperate colors to show their percentage of successful passes in that particular zone, etc.

Zones.gif
I know I've seen zone charts for the field cut into thirds both length-wise and width-wise for college (someone posted one for Mac Jones a week or two ago on Twitter that I saw). The data set I'm using doesn't have location of target so I'll have to find another set.

Unless someone else can find it.
 

JHUNSY

Well-Known Member
Aug 31, 2013
5,294
3,097
113
Des Moines, IA
I know I've seen zone charts for the field cut into thirds both length-wise and width-wise for college (someone posted one for Mac Jones a week or two ago on Twitter that I saw). The data set I'm using doesn't have location of target so I'll have to find another set.

Unless someone else can find it.
Yeah, I feel like that would really open up situational analysis. Like in third and long, who and where does Spencer tend to look, etc.

Tried googling around for any college chart data but no luck so far. Surely there has to be something out there. One problem to an alternative is most play by plays don't specify left, middle, or right to some degree otherwise I'd say maybe we could chart the data through a bit of index matching in Excel.
 

PSYclone22

Visual Analytics Mercenary
SuperFanatic
Aug 15, 2012
5,084
3,201
113
Des Moines
Yeah, I feel like that would really open up situational analysis. Like in third and long, who and where does Spencer tend to look, etc.

Tried googling around for any college chart data but no luck so far. Surely there has to be something out there. One problem to an alternative is most play by plays don't specify left, middle, or right to some degree otherwise I'd say maybe we could chart the data through a bit of index matching in Excel.
 
  • Like
Reactions: JHUNSY

Cy$

Well-Known Member
Sep 1, 2011
24,064
5,633
113
Ames
I've been messing around with some analytics using play-by-play data and thought I'd dump a few things here for people to parse through.

First: 1st down tendencies
In the first half of games OK State is calling a run on 1st and 10 nearly two-thirds of the time. Iowa State is at 57% run ratio in the first half. Not much changes for them in the second half, but we get a little bit more run heavy.

From a yards per play perspective, OK State has been pretty inefficient on 1st downs. That could be a key advantage for us on Saturday.

View attachment 76931

Quarterback:

Spencer Sanders should be back Saturday so I'm not showing anything for OK State, but here's a bit for Purdy.

On 1st downs Purdy is more efficient. And by quite a lot. If we can have a GOOD BALANCE of passing on 1st downs to stay ahead of the chains to ensure a lot of short yardage 2nd and 3rd downs or even 1st down conversions we will be successful. I know the 62 pass attempts from last year has people balking at passing, but look at the above chart: we gain >3 yards more on 1st downs by passing than rushing.

View attachment 76932

RB Comparison:

From a strict yards per carry perspective, Hall outgains Hubbard 6.0 to 4.9.

That said, they both tend to cluster their carries at around 3 yards per carry (Breece's most common yards gained are 1 yard and 2 yards, Hubbard's most common carry gets him 3 yards). The difference though is that Hall has produced a few more explosive plays for long yardage (his 3 longest carries are longer than any carry Hubbard has produced).

View attachment 76933

Receivers

The first couple tables show who each team is targeting on 1st downs. Tylan Wallace garners 50% (!) of Oklahoma State's targets on 1st downs. And there's a good reason. Everyone else has been garbage on 1st downs as receivers for them.

For Iowa State, Xavier Hutchinson leads the way with 8 targets on 1st downs, although I'd hazard to guess Kolar will overtake him here eventually. And whereas the Cowboys only get production from Wallace on 1st down passes, Purdy can spread the ball out to just about anyone and be successful.

View attachment 76935

The final charts here reflect receivers on all plays for each team, measuring completion percentage (color and label), yards per target (length of bar), and number of targets (width of bar).

Landen Akers leads Iowa State with nearly 11 yards per target. The completion percentage to him is lower, but completion percentage is less of a big deal for me than yards per target, since completion percentage can be very depth-of-target dependent.

For Oklahoma State... well, again there's no surprise here. Just cover Wallace, right?

View attachment 76936

Share your thoughts, let me know what you like / don't and even any other visuals or analytics you might be interested in. All of this is produced via play-by-play data.
Bonus material on discord
 

Latest posts

Help Support Us

Become a patron