Post Game Eval

The game is over. Hopefully we won (OBVIOUSLY). Either way, we can still use stats to evaluate certain parts of the game. In almost every game, I found stats I liked that showed things we did well and stats I didn't like that I think we could improve upon. This is happen in Wins and Losses. It is not one sided. I hope no one thinks that you will only see good stats in wins and bad stats in losses. I believe you have to look at both good and bad stats in both wins and losses. There will always be something to gain.

Similar to my Pre-Game Stats, I have my Post-Game Stats Report. As always, I will look at a lot of information but not everything goes into the Report. As a reminder, some of the resources I use for Post-Game Eval are HD Intelligence, Pivot Analysis, and my own excel spreadsheet with formulas and charts. In addition, Just Play Sports Solutions, CBBAnalytics.com, and HerHoopStats.com also provide game by game analysis that I also look at.

Here is a link to view some of the reports from this past season:

https://www.dropbox.com/sh/kbc9ew0ckjiuebe/AAAplYji8tf8wKo8ehehYta8a?dl=0

For the Post-Game Stats Report I will be discussing on our game vs Memphis. This was our 2nd time playing them this season and win #4 of our late season 7-Game win streak. The reason I mention this is as a reminder to always look at the bad stats to see what you can learn while your team is on a wining streak.

Here we go, Page #2

This page is a spreadsheet I built into Excel. Using Dean Oliver's formulas (from his book Basketball On Paper) for Individual Offensive Rating (ORTG), Individual Defensive Rating (DRTG), Usage Rate, Offensive Rebounding % (OREB%) and Opponent Offensive Rebounding % (DOR%) I am able to type in the box score from the game into a different sheet and this chart then automatically gives me these numbers. For Turnover % (TO%) I am also using Dean Olivers formula but I modified the formula to also give some credit to each player for their Assists.

This gives an overview of how each player performed on their own. It is looking at just their own individual stats (not necessarily how they impacted the team). Take note that for ORTG the higher the better and for DRTG the lower the better. For Usage Rate, in theory you want everyone to be right around 20. Why 20? because at any time there are 5 players on the court using possessions. So 100% Usage Rate is what everyone would equal. Then 100 / 5 = 20. This will not usually happen BUT this gives a baseline level of how to look at this if you have never looked at Usage Rate before. This is also how I explain it to the team, so they gave have an idea of how much they are ending possessions. The idea is, you have a high ORTG and a low Usage Rate. This means you are being extremely effective and efficient with your playing time.

The first thing I look at is the TOTALS line which is our team stats. Personally, I love to see us above 100 ORTG. BUT when your Net Rating (ORTG - DRTG) is 27+, that is a good game (in this game 92.7 - 65.3 = 27.4). No reason to be upset that you weren't over 100 in ORTG. Next I look at the Offensive Rebounding % for both teams. We have a goal of getting 40% on our OREB. We didn't hit the goal BUT we did better on the OREBs than our opponent - I consider that a good thing. Anytime you can be better than your opponent at a Four Factor stat, it is a good thing (more on this later). Then the TO% - 14.7. Pretty low. It is around 1 out of every 7 possessions did we have a TO.

Next I take a look at our player's stats. The first that stands out our #30 Tada's and #44 Jaz's numbers. GREAT ORTG, GREAT DRTG, and both below 20 Usage Rate. Without looking at the box score to see their stat lines, I know they both had extremely efficient games. Looking at the box score, neither missed a shot from the field, neither had a TO, both had an Assist. Good things all around. And with their Usage Rate being below 20 - that means, considering the amount of minutes they played, they were using possessions in a good way.

On the flip side, #1 Bria and #10 Brit did not have great numbers. Without looking at the box score, I can assume they missed shots and had turnovers. Those two things are usually big indicators for low ORTG. In this case, combined they shot 2-14 (0-4 for Bria and 2-10 for Brit) and both had 2 TOs. Remember, even in a win with a Net Rating of 27.4 as a team, individuals performances might tell a different story. As well, both of their Usage Rates we right around 20, which considering their low ORTG, is fine. If their Usage Rates were higher with low ORTG, then that is a bigger problem to look into. That means they are and us as a team, are wasting possessions. We would be getting nothing out of those. Now for #4 Tae'lor, she has a low ORTG and pretty high Usage Rate. Not what you want to see BUT numbers like that I will always take a look at the box score. In this situation, she only played 4 minutes. A low playing time will definitely have an impact on your numbers. This is a situation where you need more context.

I didn't talk too much about the DRTG. These numbers are impacted by each player's Steals, Blocks, DREBs, and Fouls. In addition, I changed the formula a bit to include our Intangible Stats. I found a way quantify stats we find important on defense and include them into their DRTGs. Which is Page #3:

These are all the stats we track for each game. Each is assigned a value. The value comes from each coach giving a value between 1 to 3 and -1 to -3, then I averaged the values out. This way we could balance each stat semi-evenly amongst the staff. From here I would be able to calculate each players Score/Possession to see how they were impacting the game on a non-box score level. Most of these stats are on the defensive end, since those numbers are bit harder to quantify. Keep in mind a lot of these stats are subjective and different coaches had different definitions. Since I was the one cutting the film and tracking these stats, it was vital I discuss with our staff what we wanted the definition of a good/bad close-out to be, contest/no contest to be, etc.

This is by no means a perfect system but it helped. The way I looked at it and explained to the staff and team, was that the Team Score/Poss should be 1.00 or higher. That basically means we did more positive things than negative. Similar idea for each player, but it was more difficult to break 1.00. So the closer to 1.00 the better.

After this was all calculated, I then found a way to include some of these Intangible Stats into the DRTG. I only used the Defensive stats and I tried to include only stats that everyone has the same opportunities to gather stats. For instance, I didn't include Ball Pressure/No Ball Pressure since our Posts/Bigs don't have the same opportunities as our guards to gather those stats. I know this is a bit confusing. For each player's Score/Poss it includes everything on this sheet. For each player's DRTG, it won't.

Page #4

This colorful box score comes from HD Intelligence. LOTS of information on this page.

The first thing I look at is the top row with all the circles. They use green, yellow, and red circles. Green is good, Red is bad, Yellow is average. We were either good or bad in this game. I could spend a lot of time explaining and talking about their information. I am going to pin point a few things that stuck out from this game.

This sheet is page 1/3 that HDI provides. This is our offense. First our PPP of 0.97 is good. It is above average and super close to 1.00. Next directly to the left of the 0.97 green box are shot distribution from 3s, 2s, and At the Rim (RS) and shot % for those three areas. This gives a super quick but informative idea of how you shot from each area. Stat wise, you want high numbers for RS and 3s and low for 2s. In this game, we did great - 51% from RS, 42% from 3s and only 7% from 2. But the shooting % was a bit different. We did good for our RS at 61% and 50% for our 2s but only 19% from 3 - something to keep note of)

Next I look at the Offensive Efficient Analysis chart. First I look at our PPP in Halfcourt and Transition. Our Transition this game was a bit low. For us, we usually have a high Transition PPP - something to look at the film on to see what we were or weren't doing as efficient there. But something HDI has on there (and it ties into our poor 3% shooting) we went 2/11 from 3s in Transition. This helps explain our low Transition PPP.

Next I look at our 1st half vs 2nd half PPP. Notice right away a HUGE difference. 1.05 PPP in the 1st half vs 0.85 PPP in the 2nd - a 0.2 PPP difference. (If you don't like decimals, think of it as thought we had 100 possessions in each half then those numbers would be 105 to 85, a 20 point difference. We were 20 points better in the 1st half than in the 2nd). This will definitely turn into something we talk about amongst the staff. That is too big of a drop off. There are a variety of things we would need to look at to gain a better idea (film, lineups, etc). but if you look at the chart, a big point of notice was our OREB% dropped from 27% in the 1st to just 8% in the 2nd. That alone tells a story.

Next on Page #5 we look at our Defensive numbers:

Similar to our Offensive Page, I could spend a lot of time talking about and explaining these numbers.

I will focus on the same points I did on offense. First we held them to 0.67 PPP for the game. Great number! Super low. Always want to see that. Next their shot distribution. As for defense you want lower numbers for RS and 3s and a high number for 2s. We didn't do a great job at limiting their high value shots. We allowed them to get 43% of their shots at the rim and 35% from 3 - some high numbers. And only forced them in to 22% of their shots from 2. BUT we did a good job at defending all the areas and held them to below average shooting %. 45% at RS, 28% from 3 and 18% from 2. We would win a lot of games holding every opponent to that type of shooting.

Next I would take a look at how we did in Halfcourt and Transition. Halfcourt was great, holding them to 0.56 PPP. Transition was, what I would consider, ok. 0.77 isn't super high but it would be great to see that be a bit lower. We just gave up too much at the rim in both areas. A positive stat is that we forced them in a TO in 23% of their Halfcourt possessions - almost 1 out of every 4 possessions, we forced a TO.

Now the 1st half vs 2nd half. 0.58 to 0.80. That is a 22 point increase. Way too high in going between halves. We gave up more shots at the rim, they shot better, didn't force as many 2s, and didn't force them in TOs as much. In addition, looking at just the 2nd half numbers on Offense and Defense, we were 0.85 and they were 0.80. Maybe it was competitive but that is just too close with a too big of increase from 1st to 2nd half. All this considering our 1st half was 1.05 vs 0.58 - a 47 point difference in the 1st half! To only a 5 point difference in the 2nd half. (For the game, the final score was 71-50, we outscore them 40-22 in the 1st half and only 31-28 in the 2nd half.)

Page #6

This is a Shot Chart HDI provides for each game. The numbers at the top are the same from the previous 2 pages. I enjoy this graphic because it a good visual that I can show the team pretty fast to show them the difference in our high value shots we took and the low value shots we forced the opponent into.

Page #7 - is a write up that HDI provides highlighting the previous 3 pages. Always brings more insightful information that I may have missed.

Page #8

This is our Potential Assists. I wish I could take credit for this but last off-season I heard a couple speak on this and have read a few articles over the years about how to get better idea of Assist numbers could be since the person shooting the ball has to make the shot to even get an assist. I decided to see what our numbers looked like this year.

A quick break down:

Assist - actual assists

FT Assist - any time a player was fouled while shooting AND if they had made the shot it would have been an assist

Potential Assist - would have been an assist if the player they passed to made the shot.

True Assist - Assist + FT Assist + Potential Assist

UA FGM - Un-Assisted Field Goal MADE

UA FGm - Un-Assisted Field Goal MISS

UA Total - UA FGM+UA FGm

I will cut the game in SportsCode to get all the data. Then I export the CSV and paste into my spreadsheet that it will then automatically calculate the above numbers based off some formulas I have in each cell. The formulas read the data that I just pasted.

If you look at the chart on the right, you can see %'s pretty quickly. The first thing I look at is True Assist % and UnAssisted %. For this game 70.3% of our possessions were True Assist and only 29.7% were UnAssited. Based off that alone, we had pretty good ball movement. Huge take away from this game.

Looking at the PPP:

True Assists - 1.02

Assists - 2.33

FT Assists - 1.33

UnAssisted - 0.82

UnAssisted FGM - 2.00

(Potential Assisted and UA FGm will always be 0.00 PPP since they are missed shots).

In my opinion, it is very easy to see that, in this game, anything with a form of Assist was a high value option. 102 vs 82 in True Assists vs UnAssisted.

I will say during the year, I learned I needed to change a few things for next season. For instance, I spoke briefly earlier about opportunities. Same applies here. I shouldn't count it against the team if the opportunity for a potential assist is not there. This means on breakaway transition situations, if there is no one in front of the ball handler with an easy lay-up, I should give them an UA FGM. I shouldn't count it. Same goes for Offensive Rebound putbacks. There isn't an opportunity for a potential assist since we want them to get the rebound and go back up quickly to score. I shouldn't count it. I also don't count real possessions, only possessions that end in a shot or FT to calculate these numbers. Something to keep in mind as I go forward with this stat.

Page #9

This is our segments from this game. I am not going to spend a lot of time talking about this sheet as I did a break of segments in an earlier post (https://dbrobertson09.wixsite.com/basketballthoughts/post/let-s-talk-about-stats).

Quickest take away from this is that we won the segments 4-1 in the 1st half but tied 3-3 in the 2nd half.

Page #10

This is our Estimated Wins. The formula is based off the Four Factors What this means is that if we had these same stats and help our opponents to these same stats for a whole season, we would win 15 games. A few things to note with this:

1. The formula is based off 14 games in a season (that is a little long to explain for this post). We "beat" the formula for estimated wins since we would win 15 games on a 14 game formula. I'm good at math but not great and I don't know how explain this perfectly.

2. Notice the TOV% is different than from page #2. That is because this TOV% does not account for Assists. My spreadsheet does not have the capability to account for opponents assists in their TOV%. I had to make sure all the stats were calculated the same way for Wins Estimated or it would not work properly.

3. This formula only works for my conference. I ran a regression using my conference team's stats to calculate an Estimated Wins Formula. If I switch conferences I would have to do this all again.

   1. As well, we were transition back into a "normal" season, I will calculate a new Estimated Wins to account for the most recent season. There might not be a ton of change but it should still be included to not completely outdate the formula.

I found that this really resonated with the staff and players. This was a very clear picture for them when I would say "We played like a 15 win team in a 14 game season" or "We played like a 2 win team in a 14 game season." (and yes that happened during the season). I have anther chart that breaks this down by the entire season, wins, and losses. During the season, what stood out was how we played in our wins vs losses. For this formula, you still want to play like a 6-7 win team even if you loss the game. This means you were staying competitive instead of just giving up. We had one game were we were a 7.4 win team. We won the game by 2.

It brought me a lot of joy that once we had played 14 conferences games, the formula estimated us to be at 10.1 win team and we were 10-4 in conference. For the entire conference, every team was with in 1 win of the Estimated Wins except for one team that was 2 games higher than estimated. That can happen. There is always a luck factor involved but that is a whole other topic for much later.

The last page of this report is our Team Goals. These change year to year. I am not going to spend time discussing them.

And there you have it. That is my Post-Game Report. I did these for every game this year. As always, there are a TON of stats that can be looked at. This was the information I decided to pass along to the staff. If we decided to show the players, I would do my best to highlight the biggest take aways without overloading them with information. Remember it is vital to communicate efficiently. As the season progressed, they would start asking more and more questions about what some of these stats meant and how the numbers came to be. Always great seeing them learning and wanting to grow.

Stats is something I enjoy! I am always looking for new methods and ideas. I am always happy to discuss!