Strength of Schedule Power Rankings
Posted: Fri Oct 05, 2018 7:45 pm
Wanted to wait until after Sim 2 to have some good data.
Your first question is probably: alright, so what exactly am I looking at?
So, I've taken everyone's schedule, copied it into a spreadsheet, and figured out how to reasonably format it. Next, I've inputted W's and L's from each game through the first two sims (and will continue to do so throughout the season). Then, I've created the following metrics which I think are important to understand. And lastly, I've weighted each column to create a 'Raw Power Score' which is a combination of how good a team has done, and how good it's likely to do based on the various metrics. Now, the first thing I might hear is 'so and so team is ranked above so and so team, but they just LOST to them' and to that I'd say, 'Tani, listen. Matchups are a thing, and there are no perfect rankings, but I'd like to think this comes reasonably close.' Essentially, record doesn't tell the whole story, and I'm trying to tell MORE of that story. There are things like trades, missed dcs, or things like this that could make these numbers funky, so take them with a grain of salt.
Win% - Team's current Win %
Diff - Team's current point differential
Opp Win % - This is the basic Strength of Schedule. It's the winning percentage of all the teams you've played thus far. A higher number is 'better' as it means you've played a harder schedule, so your record could theoretically be even better if you played an average schedule.
ROS SOS - The Win % of the rest of your opponents for the remainder of the season.
Difference - This is the difference between the previous two metrics, a positive number means your schedule will be easier than it has been, a negative number means it'll be tougher.
Remaining Home % - Percentage of your remaining games you'll play at home.
Away Opp Win % - The Strength of Schedule of your away games
Home Opp Win % - The Strength of Schedule of you home games
Win% vs. .500+ - Your winning percentage in games vs. teams .500 or better
Games vs. .500+ - How many games you've played vs. teams .500 or better
Raw Power Score - Proprietary score based off the weighting of the above categories (including current record, which is weighted the heaviest)
Alright, so what can we infer from this after Sim 2?
I think we can infer a few things:
1. Man is there parity here. The top 13 teams are within 2x of the winner, which is really really close with these numbers. The records tell this story as well. No winless or undefeated teams.
2. The Pacers are probably worse than their record. One of the things I look for when analyzing these numbers are winning teams toward the bottom, or sub-.500 teams toward the top. The Pacers are probably over-performing their true talent a bit. Look for them to course correct (or this can be the first paragraph of Pauly's GMOY nom).
3. The Pelicans and Cavs are probably better than their records - they've played a whopping 12 of 16 and 12 of 15 games against winning teams, respectively.
4. The race for the #1 lotto slot is ON, and it'll be a nailbiter. As of right now, there's no dominatingly bad team.
Would love feedback, and up for any suggestions on additional metrics to track. If I can reasonably figure out how to it, I'm game! Will keep this updated throughout the season, so let's see how teams move around!
Your first question is probably: alright, so what exactly am I looking at?
So, I've taken everyone's schedule, copied it into a spreadsheet, and figured out how to reasonably format it. Next, I've inputted W's and L's from each game through the first two sims (and will continue to do so throughout the season). Then, I've created the following metrics which I think are important to understand. And lastly, I've weighted each column to create a 'Raw Power Score' which is a combination of how good a team has done, and how good it's likely to do based on the various metrics. Now, the first thing I might hear is 'so and so team is ranked above so and so team, but they just LOST to them' and to that I'd say, 'Tani, listen. Matchups are a thing, and there are no perfect rankings, but I'd like to think this comes reasonably close.' Essentially, record doesn't tell the whole story, and I'm trying to tell MORE of that story. There are things like trades, missed dcs, or things like this that could make these numbers funky, so take them with a grain of salt.
Win% - Team's current Win %
Diff - Team's current point differential
Opp Win % - This is the basic Strength of Schedule. It's the winning percentage of all the teams you've played thus far. A higher number is 'better' as it means you've played a harder schedule, so your record could theoretically be even better if you played an average schedule.
ROS SOS - The Win % of the rest of your opponents for the remainder of the season.
Difference - This is the difference between the previous two metrics, a positive number means your schedule will be easier than it has been, a negative number means it'll be tougher.
Remaining Home % - Percentage of your remaining games you'll play at home.
Away Opp Win % - The Strength of Schedule of your away games
Home Opp Win % - The Strength of Schedule of you home games
Win% vs. .500+ - Your winning percentage in games vs. teams .500 or better
Games vs. .500+ - How many games you've played vs. teams .500 or better
Raw Power Score - Proprietary score based off the weighting of the above categories (including current record, which is weighted the heaviest)
Alright, so what can we infer from this after Sim 2?
I think we can infer a few things:
1. Man is there parity here. The top 13 teams are within 2x of the winner, which is really really close with these numbers. The records tell this story as well. No winless or undefeated teams.
2. The Pacers are probably worse than their record. One of the things I look for when analyzing these numbers are winning teams toward the bottom, or sub-.500 teams toward the top. The Pacers are probably over-performing their true talent a bit. Look for them to course correct (or this can be the first paragraph of Pauly's GMOY nom).
3. The Pelicans and Cavs are probably better than their records - they've played a whopping 12 of 16 and 12 of 15 games against winning teams, respectively.
4. The race for the #1 lotto slot is ON, and it'll be a nailbiter. As of right now, there's no dominatingly bad team.
Would love feedback, and up for any suggestions on additional metrics to track. If I can reasonably figure out how to it, I'm game! Will keep this updated throughout the season, so let's see how teams move around!