Open AI Championship

Round 2

Finals: before start

Project

Sandbox

Round 1

Round 2

Finals

Recent comments

12 december 13:31: alex3d wrote comment on post Some more good news ;)
12 december 12:07: newpuk wrote comment on post Finals
12 december 09:50: leloy wrote comment on post Finals
11 december 23:48: VexfxjC wrote comment on post Some more good news ;)
11 december 23:38: alkozel wrote comment on post Some more good news ;)
11 december 18:26: Jeners wrote comment on post Some more good news ;)
11 december 16:59: VexfxjC wrote comment on post Some more good news ;)
11 december 16:57: VexfxjC wrote comment on post Some more good news ;)
11 december 16:08: OREZ wrote comment on post Some more good news ;)
11 december 11:58: Eugene713 wrote comment on post Some more good news ;)

VK Group

Telegram chat

 

Round 2 will start on December 8, 2017 at 21:00 UTC and finish on December 10, 2017 at 21:00 UTC. Best 300 strategies from Round 1 and additional 60 strategies from the Sandbox will take part in it. Format of the games will be 2x500 with facilities. This stage, same as previous, will consist of 2 parts 12 hours each, and a break in between 24 hours long, so that participants can improve their strategies. Games in each part will use latest correct strategy, submitted before the beginning of the corresponding part. Games will be held, like before, in waves. Each wave each strategy will take part in exactly one game. In the end rating will be calculated, similarly to Round 1, and 50 best strategies (in case of equal rating participant who submitted his strategy earlier has priority) will pass to the Finals to find out the best one.



Zorander

Zorander

Saint Petersburg

Interestingly, additional 60 from the sandbox will not be taken into consideration already past?

Nov 27, 2017 12:11:26 PM
WhiteAngel

WhiteAngel

Kiev

Zorander

Zorander

Saint Petersburg

Interestingly, additional 60 from the sandbox will not be taken into consideration already past?

No, of course

Nov 27, 2017 12:21:53 PM
Jeners

Jeners

Moscow

ie a game of “conditional” will be taken top strategies 360 out of the sandbox (assuming of course if any finalists who do not move)

Nov 27, 2017 1:05:19 PM
Sanda

Sanda

JetBrains

Zorander

Zorander

Saint Petersburg

Interestingly, additional 60 from the sandbox will not be taken into consideration already past?

Soon the results of the Sandbox, you should see the tab “participants, round 2” to those who transoms, not have to guess, they fall in the round or not. Offhand, it seems that 60 people transoms end somewhere in the top 250 of the sandbox.

Nov 27, 2017 1:35:07 PM
Adler

Adler

Dzyarzhynsk

Sanda

Sanda

JetBrains

Soon the results of the Sandbox, you should see the tab “participants, round 2” to those who transoms, not have to guess, they fall in the round or not. Offhand, it seems that 60 people transoms end somewhere in the top 250 of the sandbox.

http://russianaicup.ru/contest/1/standings/withoutWinnersOf/2

Nov 27, 2017 1:47:52 PM
Sanda

Sanda

JetBrains

Adler

Adler

Dzyarzhynsk

http://russianaicup.ru/contest/1/standings/withoutWinnersOf/2

It is that strange- 60-e a place in this list is now in the sandbox 370-th place. With 360 members in the second round is not very clear, how can it be (:

Nov 27, 2017 2:06:38 PM
4way

4way

Organization::Первый Бит

Sanda

Sanda

JetBrains

It is that strange- 60-e a place in this list is now in the sandbox 370-th place. With 360 members in the second round is not very clear, how can it be (:

it is necessary that the profile was the age. So participants - complete your profile

Nov 27, 2017 2:16:08 PM
Sanda

Sanda

JetBrains

4way

4way

Organization::Первый Бит

it is necessary that the profile was the age. So participants - complete your profile

Afaik, this limitation has always been for the finale. For the second round this was not required.

Nov 27, 2017 2:25:00 PM
KseniaZhagorina

KseniaZhagorina

Yekaterinburg

And in round 2 will be game of the construction, i.e. number of structures equal to 0? Whether the percentage of such games is comparable to current conditions in the sandbox?

Nov 28, 2017 10:06:53 AM
AlexKol

AlexKol

BB

KseniaZhagorina

KseniaZhagorina

Yekaterinburg

And in round 2 will be game of the construction, i.e. number of structures equal to 0? Whether the percentage of such games is comparable to current conditions in the sandbox?

Where did you read this. On the contrary to be building.

Nov 28, 2017 10:08:17 AM
GreenTea

GreenTea

Organization::ISDDesign

KseniaZhagorina

KseniaZhagorina

Yekaterinburg

And in round 2 will be game of the construction, i.e. number of structures equal to 0? Whether the percentage of such games is comparable to current conditions in the sandbox?

I think games of different types drop out with equal probability.

Nov 28, 2017 12:34:19 PM
GreenTea

GreenTea

Organization::ISDDesign

GreenTea

GreenTea

Organization::ISDDesign

I think games of different types drop out with equal probability.

It’s in the sandbox. While in the selection for the final - only structures.

Nov 28, 2017 12:35:29 PM
jurifoll

jurifoll

Moscow

Apparently there was meant that the rules of the “game structures” the number of buildings in each specific game randomly (up to 16). Potentially, their number in the “game structures” can be equal to 0. Like the rules stipulate only the upper limit, and is there a bottom? From myself I will add: whether there can be games without factories, i.e. with one command centres? If Yes, what is the probability?

Nov 28, 2017 1:55:13 PM
DOOMovoi

DOOMovoi

Dubna

jurifoll

jurifoll

Moscow

Apparently there was meant that the rules of the “game structures” the number of buildings in each specific game randomly (up to 16). Potentially, their number in the “game structures” can be equal to 0. Like the rules stipulate only the upper limit, and is there a bottom? From myself I will add: whether there can be games without factories, i.e. with one command centres? If Yes, what is the probability?

Games only CC or only with Factories possible. The probability depends on the total count of equipment ( addition of probabilities for each individual delivered )

Nov 28, 2017 2:21:51 PM
alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

Something my runner technique on plants does not appear. It seems everything is installed, the icon appears above the machinery plant and the units no. Help Board

Nov 28, 2017 4:29:51 PM
alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

Something my runner technique on plants does not appear. It seems everything is installed, the icon appears above the machinery plant and the units no. Help Board

The question is removed. Blunted…

Nov 28, 2017 4:35:56 PM
selya_

selya_

Гимназия №1

Information about the fact that in every game of round 2 will be facilities, is accurate?

Dec 6, 2017 3:14:27 PM
leloy

leloy

PSHS - EVC

Can I assume that there will always be facilities in every game in round 2?

Dec 6, 2017 3:31:05 PM
mixei4

mixei4

КГУ

selya_

selya_

Гимназия №1

Information about the fact that in every game of round 2 will be facilities, is accurate?

Games will be in the format of “constructions”, but there is little likelihood that structures in the map zero.

Dec 6, 2017 5:58:01 PM
selya_

selya_

Гимназия №1

mixei4

mixei4

КГУ

Games will be in the format of “constructions”, but there is little likelihood that structures in the map zero.

In the sandbox, as I understand it, there is this slight chance for + 50% on what type of card would be “no structures”, right?

Dec 6, 2017 5:59:56 PM
mixei4

mixei4

КГУ

leloy

leloy

PSHS - EVC

Can I assume that there will always be facilities in every game in round 2?

There is a small chance that number of facilities in a game is zero.

Dec 6, 2017 6:00:03 PM
mixei4

mixei4

КГУ

selya_

selya_

Гимназия №1

In the sandbox, as I understand it, there is this slight chance for + 50% on what type of card would be “no structures”, right?

Should be, Yes.

Dec 6, 2017 6:01:10 PM
leloy

leloy

PSHS - EVC

mixei4

mixei4

КГУ

There is a small chance that number of facilities in a game is zero.

How small is this chance? Around 20%?

Dec 6, 2017 7:07:38 PM
mixei4

mixei4

КГУ

leloy

leloy

PSHS - EVC

How small is this chance? Around 20%?

It’s much less than 20%. I don’t think exact percent was announced, but it should be tiny. I haven’t seen any game without facilities marked 2×500+ and in round 2 chances should be the same.

Dec 6, 2017 7:14:39 PM
Jeners

Jeners

Moscow

leloy

leloy

PSHS - EVC

How small is this chance? Around 20%?

very small (0.0…1-5%)

Dec 6, 2017 7:15:05 PM
peter.alexeev
leloy

leloy

PSHS - EVC

How small is this chance? Around 20%?

Based on prior AI cups, all games in Round 2 will be in “500+” format. I might be wrong, but I would assume the probability of having “number of facilities = 0” is more or less equal to the probability of the same when generated by localrunner (with facilities set to true of course). That said, I have never seen a “500+” game on site or in localrunner with 0 facilities. At the same time localrunner generates games with “only control centers” or “only factories” from time to time.

Dec 6, 2017 7:15:47 PM
VexfxjC

VexfxjC

БГТУ «Военмех»

good luck to everyone !

Dec 9, 2017 12:03:21 AM
alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

Good luck to all! While there is round and someone is not too lazy to answer, please tell me why QuickStart with Java eats so much memory, and people in C++ is negligible. Learn Java C++ did not have any experience at all. I understand that there is a garbage collector, etc., no destructors, but java QuickStart eats 160 MB. This is due to the object creation updVehicle Vehicle = new Vehicle(vehicleById.get(vehicleId), vehicleUpdate) ?

Dec 9, 2017 12:22:19 AM
JIo6cTep

JIo6cTep

Samara

That is all the same game 500 and 500+ will be in half in the second round? Interesting…

Dec 9, 2017 12:25:43 AM
newpuk

newpuk

Almaty

JIo6cTep

JIo6cTep

Samara

That is all the same game 500 and 500+ will be in half in the second round? Interesting…

At the top of this page says (quote) “The format of the games in this phase will 2x500 with buildings”

Dec 9, 2017 12:27:57 AM
mixei4

mixei4

КГУ

JIo6cTep

JIo6cTep

Samara

That is all the same game 500 and 500+ will be in half in the second round? Interesting…

Apparently, just crooked displays the format is until it is tested.

Dec 9, 2017 12:29:49 AM
JIo6cTep

JIo6cTep

Samara

newpuk

newpuk

Almaty

At the top of this page says (quote) “The format of the games in this phase will 2x500 with buildings”

I noticed, confused the type of game. In the sandbox is always properly displayed, view.

Dec 9, 2017 12:31:55 AM
SladeThe

SladeThe

Russian AI Cup Team

alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

Good luck to all! While there is round and someone is not too lazy to answer, please tell me why QuickStart with Java eats so much memory, and people in C++ is negligible. Learn Java C++ did not have any experience at all. I understand that there is a garbage collector, etc., no destructors, but java QuickStart eats 160 MB. This is due to the object creation updVehicle Vehicle = new Vehicle(vehicleById.get(vehicleId), vehicleUpdate) ?

Java from otzhiraet a lot of memory, then the system is not to ask. In fact most of it is not being used.

Dec 9, 2017 12:38:32 AM
newpuk

newpuk

Almaty

alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

Good luck to all! While there is round and someone is not too lazy to answer, please tell me why QuickStart with Java eats so much memory, and people in C++ is negligible. Learn Java C++ did not have any experience at all. I understand that there is a garbage collector, etc., no destructors, but java QuickStart eats 160 MB. This is due to the object creation updVehicle Vehicle = new Vehicle(vehicleById.get(vehicleId), vehicleUpdate) ?

Predpolozhitelno, the bulk of the memory in your Java virtual machine. Try to do an empty test strategy is likely the volume will be about the same.

Dec 9, 2017 12:40:02 AM
alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

SladeThe

SladeThe

Russian AI Cup Team

Java from otzhiraet a lot of memory, then the system is not to ask. In fact most of it is not being used.

thank you, I’m tired already optimized, well eats and eats, the dog. Everything seems perfect, but something is wrong thought. thanks again

Dec 9, 2017 12:48:10 AM
alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

SladeThe

SladeThe

Russian AI Cup Team

Java from otzhiraet a lot of memory, then the system is not to ask. In fact most of it is not being used.

Slade, you’re cool, put it in the right direction. Need help with the stream or a lead at least. The stream to use can be one, as I understand it. if you have min, count, etc. then again it forming. whether it is clear. but I think it is very resursnozatratno. How to be?

Dec 9, 2017 1:02:45 AM
alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

Slade, you’re cool, put it in the right direction. Need help with the stream or a lead at least. The stream to use can be one, as I understand it. if you have min, count, etc. then again it forming. whether it is clear. but I think it is very resursnozatratno. How to be?

Didn’t know how to form, well, here’s an example from StartGuy double x = streamVehicles(Ownership.ALLY).mapToDouble(Vehicle::getX).average().orElse(Double.NaN); double y = streamVehicles(Ownership.ALLY).mapToDouble(Vehicle::getY).average().orElse(Double.NaN);

Is it possible to directly get the output from both coordinates in a single thread and will it be faster and more correctly than described in the Quick start?

Dec 9, 2017 1:15:39 AM
Jeners

Jeners

Moscow

In the rules it says that round two will only soorujeniya -_-, why, then my second fight just 500 instead of +500?

Dec 9, 2017 1:24:28 AM
mixei4

mixei4

КГУ

Jeners

Jeners

Moscow

In the rules it says that round two will only soorujeniya -_-, why, then my second fight just 500 instead of +500?

Don’t believe your eyes until you prostiraetsya.

Dec 9, 2017 1:30:07 AM
SladeThe

SladeThe

Russian AI Cup Team

alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

Slade, you’re cool, put it in the right direction. Need help with the stream or a lead at least. The stream to use can be one, as I understand it. if you have min, count, etc. then again it forming. whether it is clear. but I think it is very resursnozatratno. How to be?

The stream to use can be one, there’s no way to get around. Usually the streams are used where you want to compactly write the logic and not much rests on the performance. If so, it can be 2 times the same stream to create. If it isn’t, but still want to use the stream, you can do all the heavy operations with stream (large sample, filter, sort) and then to transform the stream into a collection, for example .collect(Collectors.toList()). Then it is possible to take several streams to different operations.

Of course, all you need for the task to watch. Good old loops over a collection or array has not been canceled.

Dec 9, 2017 1:31:30 AM
kuzmalex

kuzmalex

pekach dev

Just one game for a half hour? alkozel, use the normal cycle, the stream is no good here

Dec 9, 2017 1:33:52 AM
Jeners

Jeners

Moscow

mixei4

mixei4

КГУ

Don’t believe your eyes until you prostiraetsya.

Just a kind of glitch? well, Todd frets ^_^

Dec 9, 2017 1:36:09 AM
SladeThe

SladeThe

Russian AI Cup Team

alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

Didn’t know how to form, well, here’s an example from StartGuy double x = streamVehicles(Ownership.ALLY).mapToDouble(Vehicle::getX).average().orElse(Double.NaN); double y = streamVehicles(Ownership.ALLY).mapToDouble(Vehicle::getY).average().orElse(Double.NaN); Is it possible to directly get the output from both coordinates in a single thread and will it be faster and more correctly than described in the Quick start?

QuickStart is very fast, so the performance in it at all clogged. You can try first streamVehicles(Ownership.ALLY) save to the collection and check whether faster. But if you want to perform this operation as quickly as possible, then the streams need to get rid of.

Dec 9, 2017 1:36:24 AM
mixei4

mixei4

КГУ

kuzmalex

kuzmalex

pekach dev

Just one game for a half hour? alkozel, use the normal cycle, the stream is no good here

The first full tested hour. Not a lot Yes. If it goes, the time of sending strategy can play a big role :)

Dec 9, 2017 1:36:45 AM
alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

SladeThe

SladeThe

Russian AI Cup Team

The stream to use can be one, there’s no way to get around. Usually the streams are used where you want to compactly write the logic and not much rests on the performance. If so, it can be 2 times the same stream to create. If it isn’t, but still want to use the stream, you can do all the heavy operations with stream (large sample, filter, sort) and then to transform the stream into a collection, for example …

thank you. I’m generally a perfectionist, and mostly worked with a DBMS, I as the developer bases the critical speed of access to information. Worked before on the Cache , all sorts of crap and even 1C. In the DBMS I of course strong, but are constantly arising plugging with yourself. Knowledge is not enough how to raise productivity is not yet clear, that there are such questions. Thanks for the replies, I’ll learn

Dec 9, 2017 1:40:43 AM
SladeThe

SladeThe

Russian AI Cup Team

mixei4

mixei4

КГУ

The first full tested hour. Not a lot Yes. If it goes, the time of sending strategy can play a big role :)

There are problems with the system. Now I understand. In General, capacity is enough to test much faster.

Dec 9, 2017 1:44:20 AM
alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

SladeThe

SladeThe

Russian AI Cup Team

QuickStart is very fast, so the performance in it at all clogged. You can try first streamVehicles(Ownership.ALLY) save to the collection and check whether faster. But if you want to perform this operation as quickly as possible, then the streams need to get rid of.

that’s the question actually was if I didn’t show the stream in the quickstart, that would be the norm. and since people were using them shamelessly and shipped fuck all. I believe code QuickStart scared away many this time, and the second was ispolzovali suboptimal. Hence all the brakes. should have left the first silly example and the competition would more cheerful

Dec 9, 2017 1:51:59 AM
kuzmalex

kuzmalex

pekach dev

alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

that’s the question actually was if I didn’t show the stream in the quickstart, that would be the norm. and since people were using them shamelessly and shipped fuck all. I believe code QuickStart scared away many this time, and the second was ispolzovali suboptimal. Hence all the brakes. should have left the first silly example and the competition would more cheerful

Hardly quickly guilty of something.

Dec 9, 2017 5:11:42 AM
kuzmalex

kuzmalex

pekach dev

With so many waves, too much will depend on luck.

Dec 9, 2017 5:12:49 AM
alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

kuzmalex

kuzmalex

pekach dev

With so many waves, too much will depend on luck.

Well, Yes. knee wrote the seizure of buildings and the production of a single line, listening to Chester in the end the whole world went against me, the wife wants to sleep, the daughter said, not the mill programmer, 5th brandy went 0.225 D . the cat was gone and came back. I love people and live no matter ninachto. we are the best)))))))

Dec 9, 2017 5:47:28 AM
newpuk

newpuk

Almaty

SladeThe

SladeThe

Russian AI Cup Team

There are problems with the system. Now I understand. In General, capacity is enough to test much faster.

It is noteworthy that in the first stage of the second round played the same number of waves as in the first stage of the first round, although participants became three times less. Coincidence?

Dec 9, 2017 12:18:26 PM
cheeser

cheeser

Glowbyte

alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

Well, Yes. knee wrote the seizure of buildings and the production of a single line, listening to Chester in the end the whole world went against me, the wife wants to sleep, the daughter said, not the mill programmer, 5th brandy went 0.225 D . the cat was gone and came back. I love people and live no matter ninachto. we are the best)))))))

like :)

Dec 9, 2017 12:42:51 PM
bypeso

bypeso

БНТУ

SladeThe

SladeThe

Russian AI Cup Team

There are problems with the system. Now I understand. In General, capacity is enough to test much faster.

once the opportunities are there plans to speed this up? or how it goes?

Dec 9, 2017 12:53:55 PM
selya_

selya_

Гимназия №1

alkozel

alkozel

Калужский филиал ГУ СДА при Спецстрое России

that’s the question actually was if I didn’t show the stream in the quickstart, that would be the norm. and since people were using them shamelessly and shipped fuck all. I believe code QuickStart scared away many this time, and the second was ispolzovali suboptimal. Hence all the brakes. should have left the first silly example and the competition would more cheerful

The variant with the stream, as for me, quite good for the high-level language Java. Of course, that there are additional levels of abstraction, lambda functions, etc., so that takes code. But even compared to arithmetic operations stream quite fast. I, for example, the whole code is built on streams and lambdas. As for reusing many of the stream - it is possible to assemble a collection as already suggested. But there are still options. For example, in the same scope to create an array of two double and one AtomicInteger. And do something like stream.forEach(v -> {arr[0] += v.getX(); arr[1] += v.getY(); count.getAndUpdate(i -> i + 1)}). After passage through the stream to separate the two components of an array variable count. It’s not an elegant solution, but that rotten programming - you are free to choose how you write the code. Your own style, optimize where you need it. The readability of the code where it is possible to neglect the optimization. In this spirit)

Dec 9, 2017 1:41:08 PM
Adler

Adler

Dzyarzhynsk

prompt, where I can read the angry comments tag about number of games?

Dec 9, 2017 1:42:20 PM
selya_

selya_

Гимназия №1

Adler

Adler

Dzyarzhynsk

prompt, where I can read the angry comments tag about number of games?

I am ready to write) But who is to blame for the fact that someone else will carry on the Rand, and some don’t?(

Dec 9, 2017 1:46:20 PM
Gradient

Gradient

АО "Финам"

Due to the small number of games is more profitable not to fix minor bugs. And then one commit is automatically lowered myself for 10.

It’s funny, in the first round, I too was ready to rant and rave because of the curve of randomness and small number of games, because games with toporama lowered me below a plinth. And now lucky is only two games out of 19 with fellow top-50, with the result that I unfairly pulled up.. That’ll fix the system, and go back over the first hundred))

Dec 9, 2017 2:10:03 PM
VexfxjC

VexfxjC

БГТУ «Военмех»

selya_

selya_

Гимназия №1

I am ready to write) But who is to blame for the fact that someone else will carry on the Rand, and some don’t?(

random… always random… 360 participants have at least a hundred games, make sure the place was as… ideally, each and every…

Dec 9, 2017 2:12:09 PM
kuzmalex

kuzmalex

pekach dev

http://skrinshoter.ru/i/091217/KVjRipo5.png Why didn’t it impossible to play with one and the same person???

Dec 9, 2017 2:16:49 PM
tyamgin

tyamgin

Simferopol

Adler

Adler

Dzyarzhynsk

prompt, where I can read the angry comments tag about number of games?

I created

Dec 9, 2017 2:27:26 PM
VexfxjC

VexfxjC

БГТУ «Военмех»

tyamgin

tyamgin

Simferopol

I created

you have nothing to fear ))) but I generally the opponent with the fallen strategy got

Dec 9, 2017 3:08:26 PM
Dimon777

Dimon777

Tula

http://russianaicup.ru/game/view/216362 it turns out the game in round 2 of the go and no buildings?

Dec 10, 2017 12:38:57 PM
Dimon777

Dimon777

Tula

Dimon777

Dimon777

Tula

http://russianaicup.ru/game/view/216362 it turns out the game in round 2 of the go and no buildings?

the display campaign is lame, in the beginning of the test shows that a game without buildings, after the test - that they

Dec 10, 2017 12:45:54 PM
Jeners

Jeners

Moscow

VexfxjC

VexfxjC

БГТУ «Военмех»

random… always random… 360 participants have at least a hundred games, make sure the place was as… ideally, each and every…

Support technically the statistics obtained is not truthful. And more and more relies on random than on real assessment strategy. Of course such a calculation would require more time, but it would be fair

Dec 10, 2017 12:47:18 PM
MikeWazowski

MikeWazowski

Moscow

I wonder what’s the deeper meaning of repeat games between the same opponents ?

Dec 10, 2017 10:56:24 PM
malinovsky239

malinovsky239

Saint Petersburg

MikeWazowski

MikeWazowski

Moscow

I wonder what’s the deeper meaning of repeat games between the same opponents ?

For example, different conditions (location and types of buildings, weather and terrain) can lead to different outcomes: http://russianaicup.ru/game/view/222337 http://russianaicup.ru/game/view/222478 (and these two games the system and is created in a row)

Dec 10, 2017 11:25:04 PM
kuzmalex

kuzmalex

pekach dev

malinovsky239

malinovsky239

Saint Petersburg

For example, different conditions (location and types of buildings, weather and terrain) can lead to different outcomes: http://russianaicup.ru/game/view/222337 http://russianaicup.ru/game/view/222478 (and these two games the system and is created in a row)

a weak argument

Dec 10, 2017 11:38:57 PM
malinovsky239

malinovsky239

Saint Petersburg

kuzmalex

kuzmalex

pekach dev

a weak argument

I’m not trying to prove that the repetition of the opponents is correct. Just if we assume that the system with a random draw generally has a right to exist, then there’s nothing criminal in the repetition of opponents is certainly not. Because of the system as a whole may suffer (and probably suffered) individual strategy, which easily would have passed to the final at a fair Swiss, or more fair (though still not absolutely objective - just because of the dependence on initial conditions), but not because of the resources required round-Robin. And because of the repeat opponents of serious harm certainly will be no, whereas the opposite outcomes with equal initial conditions - food for thought for both participants of the battle. If the results are the same (and it is two defeats), imagine that these were two different strategies are more or less equal - not all the same, twice to play the same or two different?

Dec 11, 2017 12:13:46 AM
newpuk

newpuk

Almaty

malinovsky239

malinovsky239

Saint Petersburg

I’m not trying to prove that the repetition of the opponents is correct. Just if we assume that the system with a random draw generally has a right to exist, then there’s nothing criminal in the repetition of opponents is certainly not. Because of the system as a whole may suffer (and probably suffered) individual strategy, which easily would have passed to the final at a fair Swiss, or more fair (though still not absolutely objective - just because of the dependence on initial conditions), but not because of the resources required round-Robin. And because of the repeat opponents of serious harm certainly will be no, whereas the opposite outcomes with equal initial conditions - food for thought for both participants of the battle. If the results are the same (and it is two defeats), imagine that these were two different strategies are more or less equal - not all the same, twice to play the same or two different?

Consider the first sentence in Your statement is the most important. Fully support it and hope that the organizers in the future will change the system of selection of opponents accordingly.

Dec 11, 2017 1:01:34 AM