Started By
Message

Why doesn’t transitive property work?
Posted on 11/3/23 at 7:42 am
Posted on 11/3/23 at 7:42 am
Asking for Themetalreb…
ETA: Looking at the first few responses, I need to clarify. I totally get that it doesn’t work. I’m not saying it does. It’s just strange that it doesn’t.
ETA: Looking at the first few responses, I need to clarify. I totally get that it doesn’t work. I’m not saying it does. It’s just strange that it doesn’t.
This post was edited on 11/3/23 at 7:50 am
Posted on 11/3/23 at 7:44 am to themetalreb
2014 Ole Miss beat Bama by 6.
2014 Bama beat Arkansas by 1.
Transitive property said Rebs should beat Arkansas by more than 1 because they beat Bama who beat Arkansas.
Hogs beat the Rebs 30-0.
2014 Bama beat Arkansas by 1.
Transitive property said Rebs should beat Arkansas by more than 1 because they beat Bama who beat Arkansas.
Hogs beat the Rebs 30-0.
This post was edited on 11/3/23 at 7:45 am
Posted on 11/3/23 at 7:45 am to themetalreb
Humans have this weird thing about not being super consistent.
Also some strengths of teams don’t necessarily bear out against certain teams.
Also some strengths of teams don’t necessarily bear out against certain teams.
This post was edited on 11/3/23 at 7:46 am
Posted on 11/3/23 at 7:47 am to themetalreb
Florida beat Tennessee
Kentucky beat Florida
Tennessee beat Kentucky
Kentucky beat Florida
Tennessee beat Kentucky
Posted on 11/3/23 at 7:51 am to Sun God
quote:
Humans have this weird thing about not being super consistent.
Especially males that are 18-23

Posted on 11/3/23 at 7:53 am to themetalreb
Why doesn’t it work?
Because there are a thousand variables that do not remain consistent week to week
Because there are a thousand variables that do not remain consistent week to week
Posted on 11/3/23 at 7:56 am to AGGIES
But the most important variables are ALL the same
Posted on 11/3/23 at 7:56 am to SidewalkTiger
Because football is sometimes about individual matchups rather than overall record, etc.
Team A might be incredible against the run but be a mess against the pass.
If they play team B who is a run based offense that can’t throw the ball they will look great.
If they play team C who can’t run but throws the ball effectively team C will probably have a huge day even if their record is the same as team B.
It’s why any effective algorithm to predict scores has to take into account relative performance down to the run/pass level on both offense and defense to perform well.
Team A might be incredible against the run but be a mess against the pass.
If they play team B who is a run based offense that can’t throw the ball they will look great.
If they play team C who can’t run but throws the ball effectively team C will probably have a huge day even if their record is the same as team B.
It’s why any effective algorithm to predict scores has to take into account relative performance down to the run/pass level on both offense and defense to perform well.
This post was edited on 11/3/23 at 7:11 pm
Posted on 11/3/23 at 8:05 am to themetalreb
quote:
But the most important variables are ALL the same
I disagree. All of these variables vary week to week.
Individual skill matchups (at each position)
Injuries
Location
Weather
Mental state
Posted on 11/3/23 at 8:07 am to themetalreb
Each day is different. If you accomplished x number of tasks today then tomorrow you should be able to do the same but we don’t.
Joe Pendry used to say every game he coached was like the opening game of the season. You don’t know how players will respond this week.
Joe Pendry used to say every game he coached was like the opening game of the season. You don’t know how players will respond this week.
Posted on 11/3/23 at 8:24 am to themetalreb
quote:
Why doesn’t transitive property work?
The context of each game, variables from weather to player energy/prep/mistakes/successes, and matchups all matter. It is literally called a game of inches for a reason. A few plays made or missed here or there could have major impacts to the outcome and momentum.
Even two talented, evenly-matched teams can have drastically different outcomes from one game to the next against each other. Just look at the history of CFB matches in the same season or to the NFL.
The outcomes against common opponents have even less bearing as a predictor of a future match between the teams that played that common opponent.
There is a complexity and level of dynamism inherent in all human activity that makes it uncertain. While the %s may show a favorite, the teams still get a vote, so it's why they play the games.
This post was edited on 11/3/23 at 8:27 am
Posted on 11/3/23 at 8:27 am to themetalreb
You're talking about amateur athletes, not robots.
Posted on 11/3/23 at 8:28 am to AGGIES
quote:
All of these variables vary week to week.
Most vary half to half, quarter to quarter, series to series, and play to play.
Posted on 11/3/23 at 8:30 am to themetalreb
quote:
But the most important variables are ALL the same

Posted on 11/3/23 at 8:35 am to themetalreb
I asked this before and got laughed at. You will too
Posted on 11/3/23 at 8:37 am to themetalreb
- matchups
- turnovers
- home field
- game plans (coaches are human and have their good and bad games)
- game play (players are human and have their good and bad games)
- kicking game
- turnovers
- home field
- game plans (coaches are human and have their good and bad games)
- game play (players are human and have their good and bad games)
- kicking game
This post was edited on 11/3/23 at 8:50 am
Posted on 11/3/23 at 8:40 am to themetalreb
quote:
It’s just strange that it doesn’t.
It’s really not
Posted on 11/3/23 at 8:40 am to themetalreb
quote:Because any given Saturday/Sunday…
Why doesn’t transitive property work?
Popular
Back to top
