Hi!!! A->B->C... C->B->A...A->C->B ...C->A->B
Among these orders which order you would like to follow in Div. 2 contest for good rating? or what's your technique? Sometimes I try to follow last order but most of the times doesn't work :(
I didn't mention D and E cause they are out of my range now :p I should try to Discover D and E ... How you try?
I like to solve A first cause solving it in minimum times can give me a good rank.sometimes try to solve C first but it don't fail severely.
I have tried C .. but it took so many timed that my rating has gone down ...
I didn't mention D and E cause they are out of my range now :p
Secret of good rating — make them not out of your range:)
thnx Bro I_love_Tanya_Romanova ... it was a inspiration .. When you were an expert ... then what you did to go further? Can I get some guidelines to follow?
I've always been a fan of the good old A-B-C-D-E. It is somewhat safe and if you learn to solve the problems fast, it's actually a quite nice order.
Bro Enchom ... usually I also follow this order .. But I can go to at best in problem C. Actually, I see some coder solve problem C first and they got good rating .. that's why I wished to know the secret ? :)
Well as I_love_Tanya_Romanova said, it's much more important to learn and improve in order to get bette rating. The order of A-B-C might give you some kind of improvement, but I don't think it will be crucially significant.
When I first joined codeforces, I could only solve A and B in Div2. I never tried any different order than the regular one, and a few months ago I reached red :)
There's a much sacrifice and hard-work need to need to reach a destination. Mercy from GOD is also necessary. Specially in programming, it's like cricket. if you practice, u can be the king :) Wish to be a Red coder someday :)
Well you are downplaying your skill. Your first three contests you solved 3 problems in each. Either way your are inspiration to all of us who practice!
Nah, I'm not. Before doing an actual competition I did some virtual ones and practicing in which I had trouble solving C at least half of the times. I was happy when on the real competition I actually managed to do it, and I would then do like 3-4 virtual competitions (that's up to 8 hours) daily. So I improved pretty quickly.
And thanks, I believe that everyone can become red after enough time practicing :P
It depends from person to person, but then again, it is an interesting question. Perhaps we can find some patterns from statistics. I am not very good at these things, perhaps DmitriyH can help us on this. Like what is the expected position of each of the 5! order in a contest. How many users used each of the 5! orders. This could be an interesting experiment. Is there really a correlation between rating and order of solve?
forthright48 ... I also think it depends on someone's skill in solving !!! But, I think , there's an factor to decrease the points of each problem during contest . Researching them , one clue can come out :)
I think it is better to first solve A and then next solve the one which is solved the most.
from my humble point of view ... I tried all the variations just like you mentioned and I was disappointed almost every time because when I solve C (I mean pass the pretests) it just fails the system test .... but i can tell you this ... B --> A is much better than A --> B and then you can try C
I like this idea too. (B -> A -> C)
It depends at lot from person to person. Personally, when I was in Div. 2, my strategy was this: I solved very fast A and B (usually within 10-30 minutes), then I focused on C. Sometimes I solved it, sometimes I didn't (usually I didn't, at the beginning).
If I didn't solve it, I'd have an average position and my rating wouldn't be affected too much (it would increase or would decrease very little), so I'd maintain myself. If I solved it, the rating would usually increase (and sometimes I'd enter Div. 1).
Having this strategy for 2 years leaded me to my current level where, in an usual Div. 2 round, I solve A, B, C and sometimes D and/or E if I'm in a good mood.
In Div. 1 rounds, my strategy depends on score distribution and authors.
My advice is to solve the problems gradually and do upsolving after each contest. In my opinion, for Div. 2, A and B are usually very-very easy and being unable to solve them fast means that you might have big gaps. Try to fill them by practicing!
thanks for your explanatory suggestion.. I have noticed from different persons record that solving A,B,C in 1st hour of contest without (wrong submission) or (wrong submission+hack) can lead to Div 2 . But it hard to solve them within 1st hour or around 1st hour. I have solved problem C for approximately 5/6 times and when I try to do it quickly then I've got system testing failure. I think, Practicing can solve this problem.
I usually do A-B-C, but sometimes B or even A is tricky...
Today I tried the order A-B-C and Alhamdulillah could pass the system test(!!) successfully :)
nice work man :D
Which order you are planning to follow in the Good Bye Contest ? :)
After a bit thinking, I have found that, the question should be "In which order the problem can be read for optimal result" ?? Because, Suppose, I have solved problem A, B. Now its time to C and D and E . For Div 2 they are commonly algorithmic so it is possible that the algorithm for problem C is not known by me but the algo for problem D is known to me cause we learn algo one by one. So, there I can solve problem D instead of C earlier and for solving order, D comes earlier than C.
So, I think the question should be about the order of problem reading instead of solving.
In Division 1, I always try B->C->A.. I guess I would do the same ordering in Division 2...
The order you solve problems in is obviously not as important as your ability to solve these problems, but for fun, let's take a look at this problem for fun.
Let's assume that problems A, B, C are worth 500, 1000, 1500 points, respectively. Let's also say you solve A, B, C in x, y, z minutes, respectively.
Note that if you solve A, B, C at times a, b, c, you will get 500 - 2a + 1000 - 4b + 1500 - 6c = 3000 - 2a - 4b - 6c points.
If you follow the order A -> B -> C, they will be solved at times x, x + y, x + y + z, so your points value will be 3000 - 2x - 4(x + y) - 6(x + y + z) = 3000 - 12x - 10y - 6z
If you follow the order C - > B - > A, we will have A, B, C solved at times x + y + z, y + z, z, so your points value will be 3000 - 2(x + y + z) - 4(y + z) - 6z = 3000 - 2x - 6y - 12z
A -> C -> B will give points equal to 3000 - 12x - 4y - 10z
C -> A -> B will give points equal to 3000 - 6x - 4y - 12z
You can use this code to calculate which strategy is most optimal.
For example, when (x, y, z) = (5, 10, 20), A -> B -> C is the most optimal way to go.
In general, it is a well-known problem — you have to sort problems by time[i]/penalty[i], if you want to get highest score.
Maybe we can write problem about it:
Input: First line has 5 integers, each describing number of points each problem is worth. Second line has 5 integers, number of submissions he will take until he gets the problem right. Third line has 5 integers, time it will take to solve each problem.
Output: Maximum possible score
I go with CBA style. You should know Div2C is Div1A, most of time it can be solved in ten minutes by Div1 coders, you shouldn't be afraid of it!
Well I usually look at the problems first. I usually do A first but if I see B/C is easy I'll do them first. If there're any other "obvious/approachable" problem I will do them first, even if it's E (but this occurs rarely). Not all Ds and Es are hard, sometimes you just need to think a bit ;) Overall, I recommend A -> B -> C unless you're very confident about solving B/C quickly.