Some scattered thoughts (apologies in advance for the length)…
In terms of how much you leave the table with, it really depends on the dynamics of the auction. I’ve had auctions where I spent up to limit and others where I leave the table with $50 or so (rarely). Usually it’s the latter that I regret more than the former (i.e., more likely to look back at the bidding and wish that I had gone a $1 more on a player that I didn’t win). I budget $20, but sometimes I’ll eat into that if I have to go a few bucks over what I want to spend on a particular player. There will nearly always be injuries or something else that comes up that compels you to cut players, so you can free up roster space for in-season FAs when the opportunity arises.
For each player under consideration, I have two dollar figures: one is a target and the other is the absolute maximum that I’ll be for a player. The target prices sum to about $20 less than the salary cap space I have. I keep track of all this on a spreadsheet and keep track of the running “surplus” (or deficit), which I define as the difference between what I pay for a player versus what I’ve targeted him as. If there’s a break, then I’ll re-assess the target values of players taking into account my acquisitions thus far. Occasionally I’ll make an exception. For example, in a 5x5 auction last weekend, Seager was the best position player that I could afford (e.g., I didn’t have the cap room to make a player for Yelich), so I went $2 more for him than my maximum
I prefer to play with a deeper bench than some. That is, I leave the table with relatively few $1 players. Others have 10+ $1 players. It really depends on your risk tolerance. Sometimes there is no difference in expected production between a $1 and a $5 player; however, the $5 might have more upside. It’s something of a crapshoot, but I’ve had more success getting breakout performances from the $3-5 guys than the $1-2 guys. The flipside is that I have a litle less money to spend on top players. But that’s okay because often you don’t want the $40 guy who goes for $45; rather you want the two $25 players who go for under $20 each. Spending $3-5 on some of your bench players rather than $1-2 is a combination of an insurance policy and a speculative investment, both of which I think are useful if you look at building your roster as an exercise in portfolio management.
In terms of hitter/pitcher split, I think that Justin (creator of the Surplus Calculator) has analyzed Ottoneu leagues and has found 62/38 is the average mix. I’m usually around that, although I’ve gone as low as 55/45 or as high as 68/32 or thereabouts. It really depends on your keepers and where the bargains are. If you’ve got most of your surplus in hitters, then it makes sense to spend more on pitchers (or vice-a-versa). In a startup league, you want to go with te flow. Last year I had a first year auction where people overspent (according to my values) on most of the top hitters early and younger pitchers, which meant that there were some great bargains available later in the auction (e.g., I got Woodruff for $12).
I really suggest having a single dashboard that you use to keep track of everything. For example, I had a single Excel spreadsheet tab that had my budget information at the top, followed by a list of 5-8 players that were candidates for the roster spot I was targeting (e.g., OF < $30, SP < $4). And then below that I had the ATC projections and valuations for each of those players. I also had Ottoneu min, max, median, and estimated standard deviation. With those data, for each player as well as the probability that each would be available at my target price. I’d rather scroll down through a single Excel workbook tab then click back-and-forth between tabs, but your mileage may vary. In my browser, I have the Ottoneu draft window and a tab to my Roster Organizer. That’s it. Again, I want as little clutter as possible.
For example, I had a 5x5 auction last weekend where my top targets were Merrifield ($24 target, $30 max), Seager ($30 target, $35 max), Rosario ($21 target, $25 max), Segura ($12 target, $15 max), and Hosmer ($5 target, $8 max). I got them for $30, $37, $23, $5, and $7, respectively. Now because I overspent my targets on Merrifield and Seager, I had to reallocate on the fly (getting Segura cheap helped–this was partly due to a glitch with the server during the draft, so the other team that wanted him couldn’t bid, but he just rolled with it). So I just spent $2 each on SP rather than $4, $1 on a backup catcher rather than $3, etc. Overall, I’m pretty happy with the team and the planning definitely worked out.
One final suggestion. For every 6 roster spots or so, have a FLEX that you can use on a guy that you might not have originally targeted but seems like a good value. For me in the aforementioned draft, that was Carter Kieboom ($5). If he were still ROY eligible, I think that he’d still be a MLB-ready, Top 30 prospect and there weren’t a lot of prospects available because this league has a 5 man MiLB supplemental roster. So anyway, keep FLEX available and budget up to $5 or so.