The Starr Engine (=SE) started as an idea to assign a more realistic score to each player that better represented their value in comparison to other players. These scores are called "Starr Scores," named after the coach (EzekialStarr) who created the formula as a reminder: these scores are based off one man's perceptions and may not necessarily reflect WIS or your perceived values of players. It should be noted, Starr Scores do not take into account for a particular offense or defense, but all of them combined. I guess it can be assumed =Starr Engine (=SE) likes best the players who shine in any set. Here are some other things =SE can do:
1- Suggests a team's scoring distribution (Inside, Perimeter, Balanced). *currently only an administrative tool*
2- Proposes a position for the player based on their current ratings and IQ (the "Starr Position").
3- Suggests starting line-ups, here taking into consideration the abilities of the entire team to determine positions. =SE does not consider the opposing team's starters when considering yours. *currently only an administrative tool*
4-Much like the Starr Scores, =SE gives team and conference scores as well. Team scores are weighted to give more value to the players who are likely to play the most minutes. It's a fast and easy way to compare your team against the other 383 teams or against your own conference. A team score higher than the conference score would indicate the coach is helping his conference.
A - =SE now also provides a separate ranking system for the rookies, using the same SS formula. Let's face it, no rookie is going to stack up against the upper classmen, but it's fun to compare them against other rookies to get an idea of how your team will develop in the coming seasons.
B - D-Scores (defensive scores) have been added as well. You might be surprised to see the difference in rankings between the D-Score players and Starr Score players.
5 - The "Team Comparison" tool: Here you can type in the name of your team and also the team of an opponent. =SE will upload all the information for you and show you the scores. You can even simulate games between the two teams and select Home, Away or Neutral locations. *Editor's Note: Completed but because of instability issues is currently only an administrative tool*
A - When calculating simulation games, the tool takes several things into consideration:
1 - Your team score: Of course. A better score means a better chance to win.
2 - Coach Skill: Currently, =SE gives a specific number here to all SIM AI schools, regardless of the team score, but for human coaches it assigns a score based off your team score. There is no way to pull coach data from WIS for what I am doing, so instead I am taking the approach "the better the team, the better the coach." This will NOT be entirely accurate as I have seen some amazing coaches take over terrible programs and also some poor coaches take over elite programs. But it is better than WIS' tool, which does not exist. ;)
3 - Home, Away or Neutral.
4 - Tempo: This affects the total number of points scored in the game. Slow tempo games are probably slightly more likely to yield "upsets."
5 - Offensive and Defensive Scores: Actually, =SE does not evaluate the entire team in these areas. Instead it evaluates the top 10 defenders on a team and the top 4 scorers.
It's important to keep in mind low Starr Scores do not necessarily mean the player is a bad recruit. IQs play a big role in Starr Scores and so likely the elite underclassmen won't stand out in comparison to the elite upperclassmen. Here is a scale to better understand the Starr Scores for your players:
| 100+ | Legendary Territory. Extremely rare. | |||||||||||||||||||
| 95+ | Super-star territory. You might expect maybe 10 of the ~4,650 players to reach this phenomenal and rare benchmark. | |||||||||||||||||||
| 90+ | All-star territory. If you have a player 90-94,that player will add extreme value to your team. | |||||||||||||||||||
| 75+ | The typical range of starting-caliber players for elite teams. A team with all 75-90 range starters would do well in the NT. | |||||||||||||||||||
| 65+ | The 65-74 point range is where you'll find your best reserve players and/or starters that serve a role. | |||||||||||||||||||
| 50+ | The 50-64 point range is where you will find a typical reserve player. They can hold the fort while your starters rest. | |||||||||||||||||||
| 35+ | Players ranging 35-49 points will hurt your team. Where possible, try to keep their minutes low. | |||||||||||||||||||
| 0+ | This is walk-on territory. Players ranging between 0-34 points should not even play under ordinary circumstances. | |||||||||||||||||||
*Editor's Note: The above chart is intended as a guide for end-of-season Starr Scores*
6 - The "Player Builder" started as a tool that- like its namesake- allowed you to literally create a player and see the Starr Score. The idea behind it was not only to give coaches the ability to find Starr Scores for any of their players in any world, but also to test and better understand how the Starr Score formula worked. More functionality has since been added: you still have the ability to build your own players (now up to 12 at a time) but can now select a team and =SE will plug in all of the player ratings for you. It's also possible to find your team scores in worlds other than Rupp using this method (called the "Quick Roster" tool)
A - The tool now also features a tool to determine the current Starr Scores of new recruits and the projected minimum Starr Scores at the time of their graduation. This can also be used for veteran players. Last (and this only accurately works for new recruits) it assigns a letter grade for the recruit based on their current ratings, potential growth, Work Ethic and seasons remaining (inlcuding a potential redshirt season if that has been selected).
7 - Coach Scores - =SE has a complete and working formula to calculate coach scores. =SE takes RPI, SOS, post-season success, number of seasons played (and when they were played) and win/loss records into consideration when calculating scores. Coaches who have not yet completed 3 full seasons are disregarded to minimize inaccurate scores. =SE tracks each coach over their last 10 seasons, with diminishing returns in value the further back it travels. =SE is programmed to follow a coach to a new program, as well. It should be noted that the formula is not designed to find the "coach of the year" but rather to report what it believes to be the most valuable coach. So as an example, if you have a coach who rarely makes the NT but got lucky and won the NT championship that year, he would still rank behind the coach who has won the NT 3 years in a row and had a down year that season. Even if Coach A did better, =SE believes Coach B is the better coach. Currently DIII Coach Scores are available in the following worlds: Crum, Iba, Rupp. The vision is to eventually bring the scores to all worlds at the DIII level.
8 - Final 4 Predictions - In DIII Rupp, =SE has an algorithm in place to predict the final 4 and elite 8 teams for that season. Predictions come out before the 1st game of the regular season.
9 - *Coming Soon* - I am nearly finished with a NT bracket that shows all 4 divisions at once. It goes a step further and will predict the outcomes of the games as well as keep a tally of total NT games played for each conference.
10 - There are various other things on the website, such as records pages, news and updates, polls and surveys... and sometimes even games that win you a free season of WIS. And I'm not done yet...
