Welcome to week three of money balling beyond. And in this week we are actually going to go beyond the original Moneyball book and the paper of hicks and sour and do two things. First, we're going to extend the period that we cover and go run our data from 1994 up until 2015. So that gives us, a chance to see whether these effects are consistent over time and see if we really can see a step change around the time of the publication of Moneyball. The second thing we're going to do, is break down the statistics that we've been looking at on base percentage and slugging and break them into their component parts. There was something slightly odd about using slugging percentage, and on base percentage as comparator, since both statistics contain the capacity to hit the ball, it's only that on base percentage includes in addition the capacity to draw a walk. What we're going to do is look at each component separately, we're going to look at singles, doubles, triples and home runs as its separate statistics. And of course, the walks themselves that the players draw and see how the valuation of those statistics changed over time using our extended data period. Now we're going to focus on the salary regressions in the way we did last week with table three of hates this hour, and the reason for that is that that's really where we expect the action to be. That the story of Moneyball is that the skill of drawing walks was undervalued prior to the publication of Moneyball, and that this was represented part of the strategy of Billy Bean at the Oakland A's. And we want to see whether that valuation change, by looking at the salaries of the players. Now, we're also going to look at a slightly different set of players, last week and hot and sour, we were looking at all players, for whom data was available. We're going to focus here just on free agents. And the reason for that is, that the bargaining positions of rookies and arbitration eligible players is so different that it could be misleading to include them in the same regression. And we know that free agent salaries are essentially negotiated in an open market. So we should, if there's a change in the valuation of skills, we should be able to detect that by looking at the salaries of the free agents. Now, just as headaches and sounded, we're going to include variables to allow for the position of the players, and we want to also allow for the experience of the players. In terms of positions will use more variables than they did will use all the positions rather than just using the two general types of positions that they defined. And in terms of experience, we want to look at, we're going to look at that in a slightly different way. What you can see here in the, diagrams is two ways of thinking, three ways of thinking about the way that experience affects, performance and salaries. So, in case one, you can imagine that with experience, salary goes up, but then as a player becomes more experienced, that rate of increase slows, and eventually very experienced players as they get to the end of their career, salaries might even go down. Then the second case involves, salaries increasing with the experience, but actually exploding after a while, the acceleration continuing as players get older. And then in the third case, you can see just a linear relationship where players salary increases with experience, but always at a constant rate. Now we can allow for those three possibilities in our aggressions, by incorporating a variable for experience, which we already calculated last week when we were constructing the data for Higgs and Sour Table three and then including the squared value of that experience variable. And, the three diagrams relate to three different possible values of the squared term. It could be negative, that's the first case positive, the second case, and zero, the 3rd case. And in our our regression model will allow us to, include all three possibilities. And we'll let the data determined for us through the coefficients that we estimate which case is really the real true one. Okay, so that's what we're going to do this week, so let's get started on in let's get started on importing the data and setting up the packages for our basic regression model.