Gillispie / coach_billyg's Dynasty Rankings FAQ / Useless Information 1) What is this? This is a website version of the Hoops Dynasty 'dyansty list' rankings I used to post to the forums from time to time as coach_billyg. It is really just a ranking of HD programs, there is a top 50 per world (per division) ranking, as well as top 100 across all worlds (per division), and some other stuff beyond that. Hoops Dynasty is a college basketball simulation game by Whatifsports.com, it is their game, trademark, and everything else. This is just a fan site, with no affiliation or endorsment with the real site / game. 2) What does this do? The dynasty rankings program grabs the history page for each team and parses out the results. Several things are calculated: Current dynasty lists - current best program rankings that rank teams for each division; works in both single world (e.g. Tarkanian) or a cross-world context. Basically takes post season success and RPI and calculates the success score for the most recent 5, 10, 20, etc seasons and then sorts accordingly, providing a quality assessment of the best programs in the game right now. All time dynasty lists - historical program rankings that rank programs across eras and worlds, for each division; works in both single world or cross world format. Takes post season success / RPI and calculates the best N year period for a program (5, 10, 20, etc seasons). Any given program only has its best N year period taken into account; if two different coaches crush it with a program, only the better showing gets ranked. The season listed on the all time dynasty rankings is the season the run started in. Best conferences list - uses the program rankings to build a conference ranking; comes in current and all time formats. Longest streaks list - calculates things such as the longest streak of sweet 16s in program history, and then finds the longest such streaks game wide. Any given program only has its best streak taken into account; if two different coaches crush it with a program, only the better showing gets ranked. 3) When does it run? The history scraping bit should kick off at midnight every Tuesday morning, and completes prior to 2am. This means as of Tuesday morning, the seasons which completes their NT Monday early morning or prior, should be reflected. Example: Monday 2:30am - Phelan season 150 rolls Tuesday 12am - the dynasty rankings program runs Tuesday 2:30am - Iba season 130 rolls Tuesday 11am - user checks dynasty rankings website. Phelan season 150 will be covered; Iba season 130 will not. 4) World strength World strength - world strength is a figure that is based on human ownership of teams. Each team in a world is evaluated in a simple way, with long human ownership by a single human being ranked the highest and total sim ownership being ranked the lowest. Roughly, a world with 50% higher human involvement will receive world strength ratings about 10% higher. World strength figures are only used in comparing teams within a single division; due to difference in number of teams and such, world strength figures between divisions are essentially meaningless (i.e. generally you don't want to compare ranking scores across divisions - one can compare all time rankings without world strength across divisions though) Where is world strength used? - in the all world current dynasty rankings; in the all time dynasty rankings with world strength; in the all world best conference rankings. It is also used but irrelevant to the single world dynasty and conference rankings (but this usage makes the ranking scores incomparable across divisions - i'll probably go back and take it out now that i have a mechanism for doing so - world strength used to be used for everything always). 5) Limitations The main limitation is that the core of the tool was written 10 years ago when there were virtually no examples of 2 coaches crushing it at the same school. Now there are examples where 2 folks have great showings at a program, perhaps 10 years apart, and the lesser showing is denied ranking. 6) Contact Contact me via sitemail to gillispie1 or via the HD forums. Or the discord. I also use the gillispie and coach_billyg IDs over the years. Useless Information: 1) How are the ranking scores calculates? basically, post season success and RPI are combined to make a flat score (technically RPI rank - like RPI 1, 2, 3), and then world strength is included which impacts things like the cross-world rankings. i've thought about converting from RPI only to something better, but in the olden days when i made this, RPI was better than NT seeding even probably, and i didn't have the schedule data to use my own ranking scheme. i basically only get the history page for each team, and use that (this started as a list of longest streaks of NT appearances and the like). what would be sweet is scraping the NT tournament page to pull the seeding, because now with the new seeding (from 5+ years ago), the seeding algorithm is of pretty high quality and would be *way* better than RPI alone. here's the points system on RPI rank and post season: //rpi 1 = 10 //rpi 2 = 9 //rpi 5 = 8 //rpi 10 = 7 //rpi 17 = 6 //rpi 26 = 5 //rpi 37 = 4 //rpi 50 = 3 //rpi 65 = 2 //rpi 100 = 1 //rpi 150 = 0 //rpi 250 = -1 //rpi 350 = -2 //rpi 450 = -3 so #1 rpi would be 10 points, #2 would be 9, #5 would be 8 (#4 would probably be 8.333, #3 8.666) for post season: //pi1 = 1 //pi2 = 1.2 //pi3 = 1.44 //pi4 = 1.73 //pi5 = 2.08 //pi6 = 2.5 //nt1 = 3 //nt2 = 4.5 //nt3 = 6 //nt4 = 9 //nt5 = 12 //nt6 = 17 //nt7 = 24 where pit1 means you made pit and lost in the first round, and so forth. nt7 means you made round 6 and won (champion). the best conferences rankings aren't quite the sum of the 12 member schools, but are fairly close to that. 2) Some ramblings about world strength and its magnitude from 2/12/21 this is the current world strength which is basically multiplied by the base score (post season and rpi). hopefully this roughly correlates to the fuller worlds because that is all it is based on. i wanted to check it for myself and it doesn't compile this nicely for me so kinda building it here in case anyone wants to poke holes in it... or is just curious which worlds are considered most competitive in the rankings? rankings across divisions are meaningless. Current world strengths, division 1 Naismith - 1.524 Wooden - 1.575 Rupp - 1.608 Smith - 1.594 Iba - 1.523 Allen - 1.587 Crum - 1.545 Tark - 1.667 Knight - 1.650 Phelan - 1.614 Current world strengths, division 2 Naismith - 1.292 Wooden - 1.256 Rupp - 1.276 Smith - 1.263 Iba - 1.411 Allen - 1.296 Crum - 1.273 Tark - 1.372 Knight - 1.403 Phelan - 1.372 Current world strengths, division 3 Naismith - 1.144 Wooden - 1.056 Rupp - 1.090 Smith - 1.152 Iba - 1.124 Allen - 1.088 Crum - 1.088 Tark - 1.081 Knight - 1.156 Phelan - 1.101 within a single division, the spread is really not that big, this isn't intended to be a huge factor. nonetheless, it destroys the ability for modern teams to get ranked in the all time because the world strengths from when worlds all had 150+ are probably 30% higher or something. still, it looks like the spread within a division right now is about 10% in each division, so the fullest world folks (highest world strength really) will be getting about a 10% higher score for the same post season and RPI, as someone in an emptier world. the fullness of the world spread is substantially higher than 10% to cause a 10% score here. i think when i was looking at pop data a couple days ago i was seeing 30% and even bigger spreads, in extreme cases. (part 2) i'm just kinda spot checking world strength, i've never been too positive that actually worked. i thought d3 wooden was supposed to be a tough one but that is probably 10 year old info, but it raised an eyebrow being the weakest. so i spot checked there. Wooden - 1.056 Knight - 1.156 those are the biggest spread in d3, that would give knight a 9.5% lead over wooden by world strength. here's the actual population data over the last 10 seasons, which is as far back as population is considered (it is not as simple as considering population, for example if a world has 150 humans most of which have been tenured at the same program longer, that would have a higher score than another world with 150 where there was a lot more churn). Human population for division 3 Knight Season: 151 humans: 93 Season: 150 humans: 92 Season: 149 humans: 78 Season: 148 humans: 85 Season: 147 humans: 93 Season: 146 humans: 103 Season: 145 humans: 107 Season: 144 humans: 113 Season: 143 humans: 135 Season: 142 humans: 137 (103.6 average over past 10) Human population for division 3 Wooden Season: 123 humans: 59 Season: 122 humans: 59 Season: 121 humans: 56 Season: 120 humans: 59 Season: 119 humans: 75 Season: 118 humans: 94 Season: 117 humans: 68 Season: 116 humans: 52 Season: 115 humans: 49 Season: 114 humans: 44 Season: 113 humans: 54 (66.9 avg over past 10) so that is a pretty massive difference. right now knight has 58% more coaches than wooden, in d3. the average over the past 10 seasons is 55% higher. now d3 has freehd which makes d3 world pop on the whole less reliable, but still - i think its a good example. so here in this case - if two guys win the title with #1 rpi - the knight guy gets 10% more credit. is it 10% harder? i don't know, perhaps, but perhaps not. but similarly, a guy with a NT2 and a 25 rpi is going to get 10% more credit in knight. is being 25th best of about 104 teams 10% harder than being 25th best of about 67 teams? probably - and probably significantly more than that. IMO, 10% is more likely too low a figure, than too high - but i made it that way intentionally - i intentionally placed world strength's magnitude of impact below where i felt it truly should be. anyway i doubt anyone cares but if anyone is curious how the sausage is made on the lists, that pretty much concludes it, i think.