PhataLerror 370 Posted March 23, 2017 Share Posted March 23, 2017 In my ongoing attempts to analyze data, I'm reaching a point where I would like to involve a wider scope of data, but the data collection process itself will be cost-prohibitive due to time. Does anyone here know of a source that would be willing to provide legacy starter data for all non-quarterback positions? I have data for quarterbacks extending back to 1966 (and before in some cases), but there's 21 other players named each week of every season for each team in the league in any given year. (The math on this works out to 22 players x 2 competing teams x 267 games per year. That's 176,220 records for all regular and post-season games played since the Houston Texans became the 32nd team of the National Football League.) Thanks for any help you can give in this, even if just in the form of a lead. Link to comment Share on other sites More sharing options...
kccrow 529 Posted March 23, 2017 Share Posted March 23, 2017 Might want to send an E-mail to Pro Football Reference, Football Outsiders, The Football Database, Advanced Football Analytics, Elias Sports Bureau, or the like. Chances are, you're going to have to pay for it. Link to comment Share on other sites More sharing options...
Chiefsfan1963 1,101 Posted March 23, 2017 Share Posted March 23, 2017 Data is BS for the most part. You have idiots claiming players today are better than those of the past. How can you compare them since the game changes between decades? QBs of the 60s didn't fling it all over like today, TEs were primarily blockers in the 60s and 70s. Players worked blue collar jobs before they made millions playing football. Nope stats are BS if you try and compare between decades. Link to comment Share on other sites More sharing options...
robgar 1,046 Posted March 23, 2017 Share Posted March 23, 2017 what kind of data? chances are what your looking for is out there and pretty easy to get. the only problem is the processing of, it to put it into a useful state. For example.. you can get game logs from every nfl game. So you need to scrape them. There pdfs.. so you have to convert them. then you need to analyze them for the data your looking for. then you need to put the data into a form you can retrieve it from. btw with the big push of data analytics and big data tools, it would be great if we could get our hands on real nfl datasets. where we could look at the actual effect of things that theres no real published stats for. such as number of yards a player gained before tackled. How many yards a player had to go to make a tackle. other things of interest, in the big data world.. can we do play recognition, via ML, how about given a large enough dataset.. Can we do play prediction.. Etc.. Interesting questions... RG Link to comment Share on other sites More sharing options...
robgar 1,046 Posted March 23, 2017 Share Posted March 23, 2017 Data is BS for the most part. You have idiots claiming players today are better than those of the past. How can you compare them since the game changes between decades? QBs of the 60s didn't fling it all over like today, TEs were primarily blockers in the 60s and 70s. Players worked blue collar jobs before they made millions playing football. Nope stats are BS if you try and compare between decades. Data, is not statistics. Statistics are compiled from data, and no not every stat is meaningful. RG Link to comment Share on other sites More sharing options...
PhataLerror 370 Posted March 24, 2017 Author Share Posted March 24, 2017 Might want to send an E-mail to Pro Football Reference, Football Outsiders, The Football Database, Advanced Football Analytics, Elias Sports Bureau, or the like. Chances are, you're going to have to pay for it. It can't hurt to try. I hope they could settle for bartering. Link to comment Share on other sites More sharing options...
PhataLerror 370 Posted March 24, 2017 Author Share Posted March 24, 2017 what kind of data? chances are what your looking for is out there and pretty easy to get. the only problem is the processing of, it to put it into a useful state. For example.. you can get game logs from every nfl game. So you need to scrape them. There pdfs.. so you have to convert them. then you need to analyze them for the data your looking for. then you need to put the data into a form you can retrieve it from. btw with the big push of data analytics and big data tools, it would be great if we could get our hands on real nfl datasets. where we could look at the actual effect of things that theres no real published stats for. such as number of yards a player gained before tackled. How many yards a player had to go to make a tackle. other things of interest, in the big data world.. can we do play recognition, via ML, how about given a large enough dataset.. Can we do play prediction.. Etc.. Interesting questions... RG I've actually collected several years of play-by-play data, and had upgraded my data parser to what was originally going to be its final iteration. (I have decided that I will do one more revision will handle some of the exceptions I discovered in the previous iteration). That worked beautifully, but I didn't get far into picking through the data before life forced me to move on to other projects. Some data is just harder to retrieve, requiring the manipulation of menu upon menu for even minute fractions of the overall data to be gathered. There's so much useful stuff unexplored by the folks that have the data. Link to comment Share on other sites More sharing options...
PhataLerror 370 Posted March 24, 2017 Author Share Posted March 24, 2017 Data is BS for the most part. You have idiots claiming players today are better than those of the past. How can you compare them since the game changes between decades? QBs of the 60s didn't fling it all over like today, TEs were primarily blockers in the 60s and 70s. Players worked blue collar jobs before they made millions playing football. Nope stats are BS if you try and compare between decades. All comparisons ought to be maintained within their own contexts. I'm not even one to care all that much about volume numbers. What I'm really looking for is starter data (something I know pro-football-reference.com maintains). What I want to know in connection with one project is which players had the benefit of playing alongside players that stuck around the league for any length of time, especially when compared with players that didn't have much of a career at all. Career longevity isn't necessarily analogous to relevance, but it seems that players that manage lengthy careers are generally some of the better players. Link to comment Share on other sites More sharing options...
eraser 722 Posted March 28, 2017 Share Posted March 28, 2017 I suggest you get a job for the NSA. Link to comment Share on other sites More sharing options...
West 6,714 Posted March 29, 2017 Share Posted March 29, 2017 He might already have one.... w Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.