Race and Profession: Self-explanatory.

Minimum Starting Stats: These are the minimum acceptable starting values at level zero for each of the 10 stats. The placement of prime stats is AFTER the +10 profession modifiers. The default (when nothing is entered or a non-number is entered) is zero.

Minimum Starting TP's: The minimum desired starting values at level zero for training points, both physical and mental. The default is zero. The program lets you know if the minimums are possible. It first checks whether the physical minimum is possible, then whether the mental minimum is possible given the physical minimum.

Optimize for Level:The level for which to maximize training points. If 50 is entered, for instance, the program will calculate the initial stat placement which yields the most training points to level 50.

Reset Button: Clears the text fields and resets them to initial values.

Calculate Button: Begins the calculation. Once this is clicked, the browser will generally be unresponsive until the program ends.

Iterations: How many stat sets to test. Each iteration is one set of stats the program tests to see if it yields better training points than the best set so far. To be sure the program has arrived at the local maximum for training points, use at least 5,000 iterations, preferably 10,000. Since the program uses a best-path-forward algorithm, however, anything over 10,000 iterations is unlikely to yield a better stat set. By 10,000 iterations, the program has already zeroed in on the local maximum. This is especially true if you get the same result twice using 10,000 iterations.

However, I included enough digits for 999,999 iterations, so those with fast computers can try it if they wish. Be warned, though, that this many iterations will take several hours on most computers. For comparison, 5000 iterations takes a little over a minute on a 533 MHz Celeron, and about 30 seconds on a PIII 930 MHz.

Physical Weighting and Mental Weighting: The relative importance of physical training points versus mental training points. If the physical point weighting is set at 1.1, for instance, the program counts each physical training point as 1.1 points for purposes of finding the placement which yields the highest total training points. The default is 1.

Optimal Stat Placement: Where stats should be placed to yield the most training points over the selected number of levels. Note that these numbers are AFTER the +10 profession modifiers to the two prime stats. Below the stats are how many total physical, mental, and total training points this initial placement yields, as well as the starting training points for this placement.

Browser Notes: This program works in at least Internet Explorer 5.5 and above and in Netscape Navigator 7.0. It does not work in Netscape 4.7 and lower, because they do not support the list boxes used.

For anything over 1000 or so iterations, both Internet Explorer and Netscape Navigator will give a message to the effect that a script on the page is hogging your processing power. This is normal, since the program DOES hog your processing power. Click "no" to let the program keep running until it finishes however many iterations you entered.

With higher iterations (over 15000 or so, depending on your processor speed) you may get multiple messages like this, every few minutes. I do not know if there is a way to turn off these messages.

Version Notes:
1.1 Added physical weighting and mental weighting.
1.2 Improved the algorithm, using a 3-dimensional array to store stat growth and few other changes to avoid redundant calculations. Program now runs about twice as fast.
1.3 Tweaked the algorithm further, for about a 5% speed increase. Changed default iterations from 100 to 1000. Added e-mail link.
1.31 Fixed a bug; when one or more minimum starting physical stats were entered, along with minimum starting mental training points, the program sometimes erroneously reported that the mental training point minimum was not attainable.