These datasets contain information from NBA draft classes and subsequent advanced stats years. The idea is to evaluate which draft classes are better than others, and in which ways they are better (did they produce more stars in the top 10, or more solid role players at the middle or end. I posted an analysis of this data [here][1], but there's a lot more to be done. For example, my initial analysis just looks at drafts as a whole, not breaking it down by top 10, top 30, or top 60 picks. I also just analyzed drafts since 2000, but the dataset I uploaded has info all the way back to 1978. This data was scraped from baskeball-reference.com There are 2 datasets: season78, which has the fields: - Season: The NBA season data is drawn from. The later year is the Season value (e.g. 2015-2016 season is 2016) - Player: Name of the player - WS: Win Shares produced that season. draft78 has the fields: - Pick: The draft pick the player was. - Player: Name of the player - Yrs: Number of years the player played in the NBA - Draft: Year of the draft. [1]: https://amidatasci.wordpress.com/2016/08/04/using-cohort-analysis-to-measure-nba-draft-value/