TOPIC:

As Tobus said, not only time and population counts for score calculation.
Obviously, time and population are the most important factor but, among some other things, counts smaller pieces like: company value, performance, number of vehicles, relation between company value, goal and time played, etc. And can be checked on game details that the points difference was because of the "small pieces".
By nature, faster games give more points, but it's wrong to think that will happen in all situations because that's joining all the pieces (some above) that will give you the final points "cake". And even the rounding can interfere a bit, ofc we won't give decimal points ^^

As an example: Sometimes, a late start means you can grab some industries/towns already developed by other players and, possibly, reaching the goal faster, not meaning that played harder or more efficiently. It's a very complicated formula that only grabs data and calculates, not knowing what really happens in the games, but we always try to make it the most accurate possible

Tips to raise company value:

Pay back your loan!!!
Transport different cargo. If you only go for transporting wood your company value wont go up.
Make sure you have active stations. So there needs to stop atleast one vehicle every month and load or unload.
Your vehicles must make a profit. With road vehicles it's hard but make long tracks and you won't have a problem with that.

There are some more but you can check it in the game, under the trophy is a button that let's you see the detailed company performance information.

@ST:
Thanks for the reply. Whatever the formula is, it seems wrong to me that a player can get a "perfect" score (assuming 9 is the highest possible) while turning in a very poor time - probably not even in the top 1000 for the year - while a player who finishes in what will be one of the fastest times of the year gets the equivalent of a "C" (6/9 = 66.66%), as do many people in between. It seems like a bug or flaw in the scoring mechanism to me. The goal of the server is to build a city to a certain size as fast as you can - the people best at this should get the best scores no? Staring late might well be an unfair advantage, but I know many of the 40-50min games were started on day 1 and didn't get 9s.

PS. If rounding is really an issue, then just round at the last step - keep all the intermediary calcs as floats until the very last assignment, then you're only rounding once and it will alway be within 1 point of the "real" score.

@Trabzz:
Yes, those are great ways to build Company Value, but not necessarily great ways to build a city. It's a City Builder server, so a bit weird to be rewarded for spending half an hour building up company value instead of building up city size?

It's not a big deal for me personally, I play to beat my personal best times and the points are just a side effect, but it does seem like the scoring could be tweaked to better reflect the goal IMO. Perhaps it's time for a review/reworking of the mechanism?

@ Tobus:
With all the variables that I spoke, I invite you to create then a "perfect" formula.
And let me correct some wrong assumptions you have: the goal is a target population, there's no place saying that getting faster will get more points! Ofc, the idea is get that goal before the competition!
When players play the game only with the intention of getting better times - and sometimes breaking server basic rules - it's because only want to showoff or get more points, not because enjoy playing OpenTTD. The concept of Community points is to give a "small" reward to registered players.
Since the introduction of "auto advertising" in our client, CV's have dropped a lot in companies doing CB games (compared among players that use it).
And remember that we're talking a difference between 6 and 9 points (check other servers points range ^^): I challenge you make a formula to work on CV, CB, BusyBee, Cargo, Rating, Cash and Income servers, prepared for the variation between short servers (e.g.: #3 and #13/#13A) and long servers (e.g.: #XX-77/K1) . It's a huge range of servers we have and points formula must work for all at same time. It's not practical to have a formula per server! So, is the challenge up?

PS: Trabzz gave good tips, but mostly for company rating (aka performance), that most of players don't know and simply ignore as part of the game basic mechanics. Sometimes players forget that it's OpenTTD they are playing, and there's some important variables taken care - doesn't matter what the goal is.

Well that explains it - if the scoring mechanism is the same one used for CV and all the other game types then it's no wonder it's not always "perfect" for CB.

Is there scope for having different systems for different servers? If so I'm happy to take on the challenge of working out a better CB-specific scoring system, and to help code it if necessary... assuming of course that other CB players also think it's a good idea.

@ Tobus:
you can always join our IRC channel(s) and we can talk more about it

But, as I said, "perfect" formula doesn't exist - what we can try to do is a more accurate one and if you help... we appreciate