Access/Excel/Other

**Pokerjoe** · 10-27-10, 09:03 AM

Excel, because it was already on my computer, because I've yet to find something I wanted to do that I couldn't do with it, and because I don't feel like learning anything else now.

But that doesn't mean you shouldn't use something else. If I was starting over, I'd do what you're doing here: ask a bit first.

I will say this: Excel sat on my computers for years without my ever looking at it. I didn't really know what a spreadsheet was, much less how to use one. Now? Well, my wife says it's like a crossword puzzle for me: something I'd play with just for the fun of figuring out the answer. She might be right.

And, to everyone, without professing an opinion on what to use, I will profess an opinion that you should be using something. Just dive in and have fun.

**rfr3sh** · 10-27-10, 09:06 AM

thanks Joe
I am in the process of scraping NHL data
I am using excel because I am more familiar with it however if someone points me in the direction of something else I'm sure transferring the data will not be too big of a deal. A question related to excel that I had was , is it better to organize the teams on 1 spreadsheet with multiple workbooks, organize them on all seperate spreadsheets, or have one huge list

I am doing both right now, saving team sheets individually and also saving one big spreadsheet with different workbooks for each team

**Borat38** · 10-27-10, 01:30 PM

I am currently in the process of putting all NBA, NFL, NCAAB, NCAAF and CFL data onto a spreadsheet. Just done with NBA 2009-10. Does anyone wanna trade Excel spreadsheets? It'd make our data transcription much faster, both of us.

Also, a question: I'm not very well-versed at anything aside from creating simple mathematical formulas in Excel. Can anyone help me with what kinds of Excel tricks there are for, say, pulling out which double-digit dogs covered in a particular season? Much appreciated.

**Peregrine Stoop** · 10-27-10, 01:38 PM

vlookup and hlookup are your friends

**Indecent** · 10-27-10, 03:43 PM

Postgresql.. I can't imagine trying to build a model without using a relational database, I'm impressed that so many people can.

**rfr3sh** · 10-27-10, 04:01 PM

I decided to divert my focus to smaller. market sports. rather then going for the big 4.with my first try

**dmolition** · 10-27-10, 04:42 PM

mysql, PostgreSQL, are way better if you have a large data set, also if you know R,
rExcel is a good add in to combine both,
im starting on collecting data too, when i have enough ill pm you if you are interested in trading.

**70kgman** · 10-27-10, 05:10 PM

I always used the old school pen and paper. I have notebooks full of random crap I tracked over the years. I am going to make the switch over to Open Office Calc (basically an Excel clone) this fall/winter though.

**Maverick22** · 10-27-10, 06:42 PM

mySQL and an ass of stored procedures automatically populate information. With a slew of java code behind the wheel...

Like indecent said. Props to you excel guys, I cant imagine using that. Seems like alot of work

**gameday10** · 10-27-10, 10:45 PM

Originally posted by 70kgman

I always used the old school pen and paper. I have notebooks full of random crap I tracked over the years. I am going to make the switch over to Open Office Calc (basically an Excel clone) this fall/winter though.

What I am doing now. Going through 09-10 nba season day by day and recording data. Thinking it might take me about a week more.

**Peregrine Stoop** · 10-28-10, 01:06 AM

pen & paper = wtf?

I write down thoughts and observations, but data collection....jeez

**Wrecktangle** · 10-28-10, 04:33 AM

Believe it or not, Excel is good enough to prototype Expected Value models especially since the error trapping is so good and visual. It is possible to use it to both data store and model in but you have to chop things into chunks and link. Since it uses Visual Basic (but not on a MAC, yet) it is good enough to built decently sized models. If you are going to run a Monte Carlo Simulation, you will have a lot more issues but small "toy" models are in reach but clunky. As far as data storage goes, some version of SQL is best, but Access is usually good enough, and here again, Excel has enough power to run smaller applications and store 2D data sets (tables).

But once you get to a point to where you are doing a lot of recursion or you are ready to move out of the prototype phase, you need to go to some high level computer language like Python (an excellent intro is posted here in the Tank if you look around), or any of a number of other languages that are available that you are comfortable in using. Many are free (as is Python) and some like R (also free) which is essentially a Statistics language that can shorten the time to build stat tests and processes.

Lot of ways to go according to your knowledge level and time available. However if your level of math knowledge doesn't extend much out of add, subtract, multiply and divide you will find yourself limited and probably won't succeed.

BTW, paper and pencil = 1970s

**Richards** · 10-28-10, 07:46 AM

OpenOffice Calc, sqlite, php and a short little C program from time to time to fillin the gaps.

sqlite is awesome, full relational database, fast, tons of tools available, almost all you would ever need for a single-user database.

On the oft-discussed "what programming language is best?", I think the answer is different for everyone. If it's Turing Complete, in mathematical terms it's "as good as the next language" so it's all about what you already know and can be efficient programming in (or easy to learn if you don't know a lang), what you like, and what has some off the shelf tools for you need so you aren't writing everything from scratch.

And for 70kgman, If the what you know/what you are good at happens to be pen and paper, why not! If it's a human operating the pencil it's as Turing Complete as R or Excel a can do the same analysis......how long the said task takes, may be another question.

**rfr3sh** · 10-28-10, 09:39 AM

glad I.am majoring in math...

**Borat38** · 10-29-10, 11:16 AM

Originally posted by 70kgman

I always used the old school pen and paper. I have notebooks full of random crap I tracked over the years. I am going to make the switch over to Open Office Calc (basically an Excel clone) this fall/winter though.

Unsolicited tip: when saving OO Calc spreadsheets, save them as MS Excel spreadsheets instead of OO spreadsheets (.ods). Saving them as MS Excel will eat up more disk space, but they will load up much quicker, and you wouldn't have to worry about overly long autosaves in the middle of inputting/manipulating data. Trust me on this. You won't notice the speed difference when your spreadsheet is still empty, but once they're full of data, you will want to pull your hair out over minutes-long opening/saving files.

**Richards** · 10-29-10, 07:02 PM

Originally posted by Borat38

Unsolicited tip: when saving OO Calc spreadsheets, save them as MS Excel spreadsheets instead of OO spreadsheets (.ods). Saving them as MS Excel will eat up more disk space, but they will load up much quicker, and you wouldn't have to worry about overly long autosaves in the middle of inputting/manipulating data. Trust me on this. You won't notice the speed difference when your spreadsheet is still empty, but once they're full of data, you will want to pull your hair out over minutes-long opening/saving files.

Do you mean save your sheets in Excel format when you are using OO? OO has an impressive list of formats it can open.

What you are saying is that OpenOffice's Excel file format (converter?) is more efficient than loading the native XML format?

I would (blindly) assume that the "Office Open XML" data format for Excel 2007 and beyond probably loses this speed edge because then both OO and Excel are back to an even keel of both having to parse a bunch of XML markup to load a spreadhseet.

I'll stop with the questions now before I get this thread too far from the original post.