Unless I can think of something pressing to do, I may take a break for a few months:
http://www.mediafire.com/file/gea5k8hhp ... sR2m9j.exe(fixed computer vs computer calculations for all block negated values in the calibrators, plus fixed the lines from the .bkr ExampleRANGE and QuickRANGE values which strip the spreadsheet of formulations without needing to revert to the backup file for those files only; added an adaptation of my calibrator for ranges and stats convert .bkr for NBA Live 2003 plus fixed a few things)
This version is functionally important if you use the 'recalculation' file for '79-08 with the intent to convert to NBA Live 2005 or 06. The 3PT Frequency calculation was referencing the wrong field, the past season stats (which coincidentally serve zero purpose that I can detect in the .dbf files). So while the ratings were accurate to the player, it was the players last season stats and rookies would get a 0 rating.
This version is more user friendly with the CALIBRATOR stuff, since I rewrote the whole README.txt file. I made some grammatical errors, but it explains everything far more orderly and clearly than before, it's a big difference if you couldn't figure out how to use it. The Calibrator stuff can run from any .dbf file you use, so if you want to 'calibrate' realism settings it should extent from my project. The only bit that depends on my project is the individual comparisons box which serves no real importance.
This version is a step up in goal. I've now recalculated all the seasons with 3PT shooting to generate accurate 3PT Frequency ratings for immediate conversion into Live 2005 or 06. The next step of my project was to pre-convert at least a number of seasons (say 1979-08) for NBA Live 2005 with the relevant ExampleRANGE[2005] .bkr set. But I'm not in a hurry to get this done, if at all since it can be done there anyway and just adds to the file size.
I played a number of games to get realism settings closer to the ExampleRANGE[2005].bkr ranges for NBA Live 2005. The realism settings can be found in the ExampleRANGE file and are only relevant to the file once the ranges have been applied to the file (they are also set to be as high as 255, so you need to copy the appropriate hex line into settings.dat with xvi32). There is presently one setting I have an issue with, which is Quickness. It is set to a very minimal 20-30 range, but it may need to a greater range. I have some trouble with Charles Barkley. But it may be also be due to the way Live 2005 overcalculates the influence of height why he is playing so poorly and many tall players have it extremely easy. As a result, I included 'height' ranges to all the games ExampleRANGE and QuickRANGE files, but it does change the display height and the relative height between players. The ExampleRANGE[2005] file retains original heights.
Explanation of the CALIBRATORS operation from my new archive:
PROCEDURE
---------
These Calibrator files endeavour to take a games statistical output and adjust the realism settings to find the most sim-like settings. The main calibration file can work with .dbf files not in my project.
The first step is to play a full game of your chosen length, I recommend doing so at the game default realism settings except for fatigue, as the default game realism settings are easy to copy and paste in the SettingsCALIBRATOR[2004][2005].ods spreadsheet. After the game, save each teams player stats (general game stats not utilized).
Then, open up ConvertPLAYERSTATS[200X].bkr. Set the file link to your NBA Live 2004 or 2005 player stats folder. Apply .bkr process using BK ReplaceEm.
Open up the reformatted player stats .txt files and copy them into the appropriate places in the SettingsCALIBRATOR[2004][2005].ods in the correct sequence: human team first [cell A9], computer team second [cell A26].
(The 100 cap SettingsCALIBRATOR[2004][2005].ods file caps off all adjusted settings to the in game limit of 100; the 255 cap file goes beyond the in game limit to the hex limit of 255, which can be entered into the SETTINGS.dat file with the hex editor xvi32.)
The next step would be to open up the DataTEAMS[ALL].csv and copy the two relevant teams lines into the SettingsCALIBRATOR[2004][2005].ods in the correct sequence: human team first [cell A2], computer team second [cell A3].
Next you should enter the played at settings into the PRESENT columns [B60-B68, G47-G68, L47-L68]. You can find the default realism settings in the RECORDS tab for copy and paste.
(The next step would be set certain weightings, options and difficulty values, but I'll get to them a little further on.)
On the RECORDS tab, the uncapped adjusted realism settings are in the column B2-B56. Copy those values and paste it into the column called GAME 0 [cells I2-I56] using paste special (deselect 'paste all', then deselect all boxes except 'numbers').
The capped numbers are found either in column U2-U56 or in the hex at the bottom of the PROCESS tab. The hex at cells A75 and A89 are the present game adjustments. The hex at cells A81 and A95 are the average adjustments of the game adjustments stored in RECORDS (NOTE: the first game, GAME 0, isn't set to be counted towards the average hence it'll read #DIV/0!, and it'll take 2 more games to get an average). Copy the relevant hex.
(The smaller hex codes beneathe the others beginning with At: are the played at hex codes based on the PRESENT values entered in on the PROCESS tab. These are to help identify the location of the hex values in SETTINGS.dat for xvi32.)
Extract the xvi32 archive. With this utility, open up the NBA Live SETTINGS.dat file. Find the location of the start of the string at 0000:0035 (NBA Live 2005) or 0000:0104 (NBA Live 2004). Go to Edit/Overwrite string... select 'hex string' and paste the hex in the box. OK it and save the file. The new settings are ready to play.
On your second game, copy the played at settings into the PRESENT boxes in the PROCESS tab (before inserting in the game statistics so they don't change) from the capped column in the RECORDS tab [cells U2-U56]. Then proceed to reformat the player stats with ConvertPLAYERSTATS[200X].bkr and copy them into the relevant cells as above. Set the weightings (if necessary). And finally, copy the uncapped current adjustments from cell B2-B56 into GAME 1, 2, 3...
STATISTICAL COMPARISON BOX
--------------------------
In SettingsCALIBRATOR[2004][2005].ods at cell beginning P43 you'll notice some player comparisons, which attempts to compare players with their real life stats (100% = accurate). To do this, it needs the season specific stats from my PLAYERS.dbf/.csv files and then to insert those stats into the player stats files (notice that in the provided SettingsCALIBRATOR[2004][2005].ods file, beginning at V9 these stats are stored in addition to the game stats and extend beyond the headers, which means the player played for multiple teams that season and the first lot, the numbers used for comparison, may not automatically reflect the correct team).
In the CALIBRATORS directory, there is a file called ConvertDBFtoBKR[200X].bkr. Apply this to one of my PLAYERS.csv files. Retitle the result ConvertPLAYERSTATS[200X].bkr. The currently provided ConvertPLAYERSTATS[2005].bkr file is preset to the 1979-80 season and will produce meaningless results for other seasons, but have no functional effect. This step is unimportant and not utilized in any formulation.
OPTIONS
-------
To get the right output, some of these settings are essential.
This spreadsheet does not attempt to take a games stats and adjust it to a real life game stats as might be assumed. What it instead does is ratios various values to other values to attempt to get the correct game stats. For example, the correct ratio of offensive rebounds to defensive rebounds, not the correct amount of offensive or defensive rebounds on average in a full game. This way, if say a human player shot twice as many shots and was blocked twice as frequently as expected, rather than having one process slow the game down to half speed and another halving amount of blocks what this does is determines the amount of blocks was accurate to the amount of shots and separates it from the issue of the games pace.
With the ACCELERATED option, if set to 0 the game pace is set to per minute expectancies. If you play four 6 minute quarters, you can expect half the stats of an accurate 48 minute game. Four 3 minute quarters, 1/4th the amount of stats. If you play an overtime period, it should set the values consistent with a 48 minute game with overtime periods. However, if set to 1, it does what it suggests: it accelerates gameplay. If you play at four 6 minute quarters, it will double the pace to try to generate an accurate accelerated 48 minute game. This presents a problem with overtime periods and if such a scenario occurs, it is recommended you save the stats for processing before playing OT, as it will either accelerate gameplay or decelerate (if total minutes played at is greater than 48) gameplay inconsistent with your other games. Accelerated gameplay is sadly not properly given with the NBA Live games, the clocks (including shot clock) should move faster for it to be ideal.
If MIDRANGE FORMAT is set to 1, mid range shooting is adjusted to attempt to get FG% to accurate levels. This is despite the amount of 3 point FGs a player attempts. 90% of the game could be shot from behind the 3PT line, and the overall teams shooting percentage might be around 33%, for example. Let's say the teams average FG% is 45%, with this option set to 1, it would increase midrange shooting to try and get the average up tp 45% despite how many 3 pointers were shot. This could conceivably have the effect of encouraging players to shoot more midrange shots, but is also unrealistic. 0 is recommended. With 0 set, it is the same as short range shots, based on 2 point FG's only.
FT FORMAT should be fairly clear: 0 sets it to take into account both teams; 1 sets it to be based on human shooting; while 2 sets it just to computer shooting.
FGA MINUS BLOCKS is for greater refinement of data regarding the amount of potential rebounds and accurate shooting percentage. An imperfection in Live 2005 (and 2004 also, I believe, perhaps even more so) is that not all blocks count as a FGA. Some block shots or ones that blur the lines between intercepts and blocks but count as blocks are not registered as FGA's despite the player going up for a shot. If it had proper statistical keeping, it'd be straight forward: 1, 100% remove all blocks from the FGA count. But if I had to guess, I'd say the number is more like 80% for NBA Live 2005, or 0.8. So this option isn't simply on/off, 1/0, but shades of grey. It is set to 1 by default.
DIFFICULTY
----------
Beginning at L44 are the difficulty settings. Neutral makes it harder for both CPU and human in regards to the realism fields that apply to both teams (FT shooting, for instance), while human difficulty divides its corresponding settings and computer difficulty multiples its corresponding the settings. Setting all to, say, 1.1 would result in both teams having a harder game on the free throw line, it will result in the human player having a 1.1 harder game in regards to the human settings and the computer player having an easier game in regards to the computer settings.
WEIGHTINGS
----------
The box beginning at A43 are the weightings split into 6: fouls, user steals, user blocks, cpu steals, cpu blocks and speed format. What I mean by weightings is that all the corresponding values impact each other. With user steals, for instance, the two values are summed up and the share of the change is appropriately proportioned between the two user steal settings. If both read 1 or are equal, they get a 50% share between them. If one reads 2 and the other 1, then the one that gets a 2 will get 66% share (2/(2+1)). If one has 0 (where possible), it gets a 0% share and is effectively switched off in the calculations, while the full change is still evenly divided amongst the other values.
All of them should be pretty much self-explanatory, except for SPEED FORMAT. There are 5 fields where the game pace adjustment can be distributed to try and effect or persuade the player to play at a lower speed to get an accurate sim. By default, it is distributed 4 ways. Game speed effects the game speed value. If all the others are set to 0 and game speed is set to 1, all of the change is carried by the game speed setting reducing the rate in which the players run up the court. If shot difficulty is set, then it makes the human shooting more difficult to try and encourage the player to take his time with his shots. If the cpu block setting is set, it makes the computer player a better shot blocker. If the cpu steals setting is set, it makes the PC better at steals and intercepts. And finally, if sim length is set, it suggests a different quarter length; turn all the rest to off (0) and then only a different sim length will be suggested. Again, all these values are summed up and divided amongst one another as shares. Sometimes there isn't enough change when it is divided up 4 ways and you may exclude ones from game to game in order to get numbers to move in a desired manner.
COMPUTER VS COMPUTER
--------------------
If you should wish to do this, you need to set computer players at B7 to 2. Then you should copy the blue row beginning A4 and paste special (text and numbers only) into the rows beginning A2 and A3. Then all human influenced speed format weightings (shot difficulty, cpu blocks and cpu steals) should be switched to off (0).