If you have been following my blog over the past 2 weeks, you should be familiar with the significant strong main effect that both Offense & Defense Points Per Game have on Wins. If not, you can read my previous two posts that independently analyzed both scoring offense and scoring defense as a main predictor of Wins. After the most recent weekend of games (October 21, 2018), I uncovered that Scoring Defense actually holds the edge in predictive strength as a Key Driver of Wins over scoring offense with a 52% predictive strength compared to 42% for scoring offense. (See Picture below)
A quick refresher on our statistical terms for this analysis
A key driver is an input field or combination of input fields that has a statistically significant effect on a target field. This effect is more than a mere association. By identifying key drivers, you can discover which field or combination of fields have the most effect on a target field.
In our analysis the Target Field is Wins and we set out to determine what the Key drivers of Wins are. As a former football coach , I am fully aware that there is nothing more important than wins & losses. With this in mind, I set out to determine what components of the game have the most significant impact on Wins. The first two key drivers that I analyzed so far have been scoring offense & scoring defense.
The other key term to understand is Predictive Strength. Predictive strength measures how well a predictor accurately predicts a target. So in this case, how well does scoring offense & scoring defense accurately predict wins? You can use predictive strength to compare the predictors. Predictive strength is displayed as a percentage. A predictor with a predictive strength of 100% perfectly predicts a target, so as noted in our analysis, scoring defense has a PS of 52% & scoring offense a PS of 42%. Because the predictive strength is not at 100%, its not a perfect predictor of wins. This is proven with teams like Georgia Tech & Missouri who are in the high category of Offensive Points Per game but have a combined record of only 7 wins & 7 Losses. Although overall, there is a higher win average amount in this high group, its no a guarantee that you will win all your games.
While Points per Game both offensively and defensively independently are strong predictors of wins, it’s not the only factor. So, knowing that each driver independently is not a perfect predictor of wins, what happens if we analyze a combination of both drivers (Scoring Offense & Scoring Defense)? Well, that’s exactly what I did!
Offense & Defense Combo Analysis: 76% Predictive Strength
I ran the predictive analysis model with a combination of both scoring offense and scoring defense as a predictor wins. The model showed that the combination of scoring offense and scoring defense drives wins with a Predictive Strength of 76%. This analysis is evidence that the combination of both Scoring Offense and Scoring Defense is a much stronger driver of wins than either driver alone. This is a lesson that teams like Texas Tech learned last year and teams like Missouri & Toledo, are learning this year. Scoring a lot points alone is no guarantee for winning, just as playing stingy defense alone is no guarantee. However, if you do both well, your chances of winning increase significantly with that predictive strength of 76%. This analysis was designed to find which combinations of scoring Offense & Defense drive wins. You can use this to see how your team matches up.
Out of 130 FBS Teams, 8 Teams stand Out
After running our analysis, we came out with multiple groupings of scoring offense & scoring defense combinations and what the average number of wins each team in that category had. In the picture below, each shaded cell represents the Win average at a combination of categories of Def Points and Sum of Off Points/G.
The graph you see above is an analysis of combined scoring offense and defense by categories along with their Average # of wins. You will notice both the Offensive Points per Game & Defensive Points per game categories that we previously analyzed are plotted here on the graph. The darker the box the higher the Average Win Total is for that combination of offensive and defensive points per game. The lighter the box, the lower the average number of Wins.
The Sweet Spot: Score more than 36.4 PPG & Give up less than 20.2 PPG
The graph has a lot of categories, so let’s examine the High & Low Groups. For example The top upper left hand box is where (Greater than) > 36.4 Offensive Points Per game & (less than) < 20.2 defensive Points per Game intersect. Out of 130 Total FBS teams, only 8 teams make it into this group. The teams in this group have an average of 6.2 wins each, an overall record of 50 wins and 7 Losses for a win percentage of 88%! This is the group you want to be in.
The Group of Death: Score less than 24.3 PPG & Give up More than 33.40 PPG
This is a scary place, the (Low Group). This group is represented in the graph as the box in the very far bottom right. This group consists of teams scoring less than < 24.30 Points per Game and giving up more than > 33.40 points per game. There are 8 teams in this group with an average of 1.4 wins each. This is certainly not where you want to be, as you can see by the list of teams below who have all struggled greatly this year. They have a combine record of 11 wins and 48 Losses.
A look at the Next Level
What I like to do at Ace Football Analytics is help coaches create realistic targets for their teams that give them the best chance at winning. It’s not about throwing out random numbers like "hey we need to score 50 ppg and pitch shutouts every week". Those are outliers and only the truly special teams can even come close to achieving that level of performance. So, after the High and Low groups, there are plenty of other categories on the graph that lead to winning football. They may not be the best of the best but you can find plenty of championship contenders and bowl teams scattered across the other categories. Lets take a look at one more combination category that drives wins.
Offensive PPG Btw. 30.7 – 36.40 and Hold Opponents under 20.2
This category holds the second highest average win total. Out of 130 FBS teams, 7 Teams fall into this category. This group is represented on the graph in the box that is second from the top on the left hand side. The teams in this group have an average of 6.1 wins each and a combine record of 43 wins - 7 Losses. Not bad at all!
Want to see the rest of the teams by each category?
For the complete breakdown of teams by category, contact me by email or subscribe for free. Stay tuned for more analysis and insights each week as we explore more combination categories that drive wins. If you have any areas that you would like me to analyze, feel to reach out anytime. As we test out more combination categories, we may uncover the perfect predictor of wins, you never know.
Comments