UFO Activity and scores as compared to the original

[Nevill]

There is something not quite right with UFO Activity compared to the original.

Out of curiosity I started 10 games in Europe  on a veteran difficulty level. In 9 of them I have detected no less than 3 UFOs in January, in some of them 5 or 6, and there was a guaranteed terror mission before the end of the month.

With OpenXcom I have 5 games where I only detect 1 UFO in January. In one of them it's April and I have detected only 3 UFOs total. Some of the games didn't have terror missions in January, though in these cases there was one somewhere in the early February.

Now, I get that these things are probably random and the graphs show UFO activity in other regions, but there is still a discrepancy between the original experience and OpenXcom, and not a good one. Now I can get positive scores without lifting a finger, and this never happened in the original. The negative scores you get for idleness are also lower (that is, you get better results) than they were in the original. Generally, I can see a 200-300 points difference in the first two months (say, -1100 vs -800 in January, and -100 vs +100 in February). Makes it harder to lose the game.

This leads me to believe that either aliens are far less active than they were, or something else is wrong.

I believe the point was raised in this thread before.

[Warboy1982]

i've checked and re-checked and re-checked and re-checked this. our initial terror/retaliation mission timers were slightly wrong, this is now fixed, and i average 3 detected UFOs and a terror mission in january with no action taken on my part.

[redv]

This is what i said about six month ago. https://openxcom.org/forum/index.php?topic=1617.0
If you will get more statistical data, you will see difference in UFO activity between levels of difficulty.
i.e. for the beginner level you will got less quantity of UFOs than veteran level.

DOS version has a bug when after first terror mission, level of difficulty will be dropped to beginner. https://www.ufopaedia.org/index.php?title=Difficulty_Levels
Therefore right statistics can be obtained for the windows version of Xcom.

[Nevill]

Yes, I referenced that very thread in my first post.

Anyway, it appears to be a weird fluke. I mean, I really did ran 10 games with abysmal UFO detection rate. Now I started 10 new ones and had 3 to 5 UFOs in January, as per original. Huh.

At least with Terror timings fixed it is now a guaranteed loss for 2 months of inaction. Thanks.

Was something done with penalties for not reacting to Terror Sites? I believe the penalty is now -800 as compared to -1000 in the original.

[Warboy1982]

it's mission points * 100
mission points for terror mission = 10

10*100 = 1000

and i hate to retread old ground, but ufo frequencies are NOT affected by difficulty until 14-(difficulty) months.
statistics are meaningless to me, considering the amount of randomness involved.
i've looked at the decompiled windows exe, there is no evidence to support such a claim.

[Nevill]

Alright, so what is with the monthly scores, then?

In original, when I am fully inactive, I get these.
Number of UFOs spotted in January across 3 games: 4, 5, 4 + 1 terror site in each game
Scores: -1100, -1117, -1117

The value of -1000 and a fluctuation of -100/200 in January is consistent across all games (about 20-ish now).

In Open Xcom, number of UFOs: 3, 2, 4 + 1 terror site in each game
Scores: -854, -843, -844.

Again, the number -800 and -50/100 is consistent.

Veteran difficulty in both cases.

[Warboy1982]

made a slight adjustment, here's a vanilla saved game that made it through 4 months with no action on xcom's part so you can compare the graphs (alien activity and xcom score) and what happens with the score, note that in this game, february's terror mission didn't appear until march, so one can very easily see the 400 point adjustment the council makes for you, and you can see exactly when it starts.

[redv]

and i hate to retread old ground, but ufo frequencies are NOT affected by difficulty until 14-(difficulty) months.
statistics are meaningless to me, considering the amount of randomness involved.
i've looked at the decompiled windows exe, there is no evidence to support such a claim.

Statistic analyzes randomness, but the result is not random.
For example, all physical laws obtained on the basis of statistical data processing and confirmed by statistics. If a claim is not confirmed by statistic, then the claim is false.

I have no doubt that you saw the code of Xcom. But I know, that code not always correct interpreted by human. Are you do mistakes in code too?

Therefore I suggest resolve all doubts. https://openxcom.org/forum/index.php?topic=2115.0

[Warboy1982]

so you're asking for people's saved games to gather statistical data, yet you didn't even download mine... any reason you're ignoring the evidence i put forth?

[Warboy1982]

apologies for the double post, but while testing some other things i managed to get a saved game with the exact same circumstances as the one i got in the original, the terror mission in february didn't happen until march, and i made it to april with no action on my part, here's the save, and some side-by-side screens of my score in vanilla and in openxcom. i'll grant you, the scale isn't identical on the score window (i'll fix that), but you can see the scores are almost 1:1.

[redv]

so you're asking for people's saved games to gather statistical data, yet you didn't even download mine... any reason you're ignoring the evidence i put forth?

Because your save not pass the screening

Yes, before a data will be processed, this data must pass a screening. Possible different criteria. This depend on task. Basic criteria is July-August 1999 for in-game time. Additional criteria, what I will take in account will be: player must to want to win. It easy to understand if you look at a bases position, experience of soldiers etc. Because I need to screen all testing saves.

It like as analyze of chess party. Makes sense to evaluate only positions when player want to win, but not does random moves.

[Warboy1982]

but player activity only adds more random variance into the results, at 6-7 months into the game, with players actively shooting down UFOs, the random element is far too great to form a realistic basis for comparison. doesn't all statistical study require a control/baseline? here i have two saves with as little random variance as is humanly possible.