Advice wanted

Historical what if discussions, hypothetical operations, battleship vs. battleship engagements, design your own warship, etc.
Keith Enge
Member
Posts: 138
Joined: Sat Jan 01, 2011 1:36 am

Advice wanted

Post by Keith Enge »

This question doesn't really involve hypothetical naval scenarios but none of the other forums fit it either. Since this forum has the most posts, I thought that I would add it here in hopes of getting replies.

I have recently added functionality to the battle finding capability of my naval database. The database includes 488 battles of all types, the vast majority of which are historical but with some "what ifs" (but, as will be discussed later, even the "what ifs" have an historical basis). My question is, what can I do to further improve the battle finding capability? If it were your database, what do you wish that it did?

For you to answer those questions, I have to list what the battle finding portion of the database can currently do. All of the tests listed below can be individually enabled or disabled. It can restrict the battles found to:

One or more of 16 regions. These regions were judiciously chosen. They range from as small as the English Channel to as large as the Central Pacific.

Battle types (surface, carriers vs carriers, carriers vs non-carriers, convoy, purely ASW, aircraft vs aircraft, aircraft vs ships, or other). Battles have a single primary type but can have any number of sub-types; matches of only the primary type, matches of any of a list of types, or exact matches of primary and sub-types can be specified.

Day, night, or both.

Historical or "what if". The "what ifs" have five categories. "Fizzle" means they met but no action ensued. "Search fails" means one side knew that the other was there but couldn't find them. "Missed" means that both sides were nearby but they didn't encounter each other. "Timing" means that they would have met but one side was delayed for some reason. "Rewrite" means rewriting history like adding a ship that ran aground the previous day or combining the actual force with another in the area but on a different mission.

Tactical winner (major, minor, or tie).

Strategic winner (major, minor, or tie).

Number of ships (Axis, Allied, or combined) (a range of numbers is specified).

"Power" ratio between the two sides (who is expected to win as determined by the database's power calculations) (a ratio range is specified).

Does it have an animated map (normal or carrier battle type map)?

Specific ship types from specific navies. For example, you can ask to find battles with Italian light cruisers and destroyers, British destroyer escorts but with no Italian heavy cruisers (the test can be much more elaborate than this).

Planes - number in the battle (range of numbers for the Axis, Allies or both), country, mission, wheeled/floatplane/flyingboat, biplane?, number of engines, type of plane (fighter, divebomber, night fighter, etc)

You can also specify a date; only battles after that date will be found. When a list of battles is found, it normally replaces the previous list found. However, there are four other options. It can be merged with the old list. It can list only battles common to both lists. It can list only battles unique to the new list. It can list only battles unique to the old list.

OK, that is what the database currently can do. What do you think that it should also do to be even more useful?

Thanks in advance for your input,

Keith


P.S. - You don't have to read any further; my questions have been asked. However, I thought that I would include this to show some other things that the database can do. The battle finding capability somewhat mirrors the database's ship finding capability. For example, you can have it find all Italian light cruisers built by the Ansaldo, Genoa shipyard between April 1933 and June 1934 with eight 3.9" secondary guns and whose number of torpedo tubes decreased by 50% or more in a refit between June 1941 and March 1942. Furthermore, you could further restrict it to those sunk by British and American bombs and shore batteries but with no submarine torpedo hits between September and November 1942 within 20 miles of Tobruk. Actually, this would find no ships that meet all of these criteria but you get the idea. Much more complicated tests can be made; ships have up to 157 different data items that can be tested, everything from secondary turret armor to the rate of fire of an AA gun. There are 48 possible ship fates, everything from kamikazes to diving accidents. The sinking locations can be specified by region (the same regions as for battle finding) so you might choose the Mediterranean. In the example, to get that area near Tobruk, you zoom in on a map and draw a box around the desired area.
alecsandros
Senior Member
Posts: 4349
Joined: Wed Oct 14, 2009 2:33 pm
Location: Bucharest, Romania

Re: Advice wanted

Post by alecsandros »

Hello - you've done a huge work !

I would add a weather control... storm/fog/blizzard/etc. Selectable from the "find battle" commadn...

Also, are there any submarines ?
Keith Enge
Member
Posts: 138
Joined: Sat Jan 01, 2011 1:36 am

Re: Advice wanted

Post by Keith Enge »

There are 2689 submarines in the database from 25 different navies. 110 of the 488 battles had submarines among the participants. In some of these, of course, they didn't play a combat role, maybe acting as scouts outside an enemy base. 66 of the battles had a primary battle type as convoy; another 24 were ASW battles, essentially convoy battles without a convoy (hunter-killer groups and the like).
alecsandros
Senior Member
Posts: 4349
Joined: Wed Oct 14, 2009 2:33 pm
Location: Bucharest, Romania

Re: Advice wanted

Post by alecsandros »

can you select the country ? OR political faction (ex Spanish Nationalists and Spanish Republicans during the 30s Civil War)
Keith Enge
Member
Posts: 138
Joined: Sat Jan 01, 2011 1:36 am

Re: Advice wanted

Post by Keith Enge »

alecsandros-

The country selection is done in the section where you specify "specific ships from specific navies" or the same thing for planes. I don't separate into national factions. In the case of Spain, this doesn't matter since the database is of WWII battles. I did stretch it to include some non-WWII air battles (three in the Spanish Civil War, four in the Sino-Japanese war, and two in the annexation of Czechoslovakia). From my point of view, there were no naval battles of note in those conflicts. However, one of the Sino-Japanese air battles that I mentioned was an attack on the Chinese gunboats Ning Hai and Ping Hai. Some of the attacking planes came for the carrier Kaga. In the WWII time frame, I don't split the Italians into faction after September 1943 either, even though some Italian ships fought with the Axis and others with the Allies.

On the other topic, I considered your weather suggestion. Weather had less of an effect on WWII era battles than in the age of sail but it still had some. I already deal with it on some level. In my animated battle maps, carrier battles can show areas of squalls such as those that sometimes hid carriers in the Coral Sea battle. Also, once a battle is "found", it can be looked at. Even if it doesn't have an animated map, it still has the participating ships (divided into their task forces or whatever) with their data accessible. There is also a short description of the battle and, as part of the description, weather is mentioned if it played a role. Fortunately, as part of my database, along with all of the other search engines, there is a powerful text search capability. Letting it loose on those battle descriptions, I had it look for "weather, squall, gale, cloud, rain, and visibility). The following is the list of battles that it found.

11/24/39 Bad weather allowed Gneisenau and Scharnhorst to return home undetected
2/18/40 Weather caused a German raid to be aborted
4/9/40 Both sides took weather damage when Renown met with Gneisenau and Scharnhorst
8/29/40 Bad weather helped prevent an Italian force from intercepting Operation HATS
9/29/40 Bad weather hindered Italian air reconnaissance of Operation MB 5
12/22/41 One of the reasons that a Wake Island "what if" carrier battle didn't happen was a delay for refueling in bad weather
2/1/42 Bad weather hindered assessment of the effectiveness of a carrier raid on the Marshalls and Gilberts
2/12/42 Bad weather was one of the reasons British planes did so little during Operation Cerberus
3/8/42 Bad weather had German and British forces unable to find each other
5/7/42 A band of squalls occasional hid carriers during the Coral Sea battle
7/16/42 In Operation Pinpoint, bad weather helped hide Welshman from her pursuers
9/27/42 Bad visibility let German raider Stier and freighter Stephan Hopkins get too close, resulting in their mutual destruction
10/11/42 Rain helped prevent the forest fires that the Japanese intended when planes for one sub bombed Oregon forests
3/6/43 Weather created convoy stragglers, helping the wolfpacks
4/28/43 Ditto
5/22/43 Betty bombers appeared unexpectedly out of the fog off Attu in the Aleutians and attacked Charleston and Phelps
7/22/43 Weather postponed the Kiska evacuation, preventing a "what if" battle
9/10/43 In an attack by the USAAF on Paramushio in the Kuriles, bad weather aborted many of the attacks
5/18/43 Weather cause most of the P-38s to abort from their planned attack on Ploesti, Romania
8/18/44 Weather caused stragglers in Japanese convoy HI-71, leading to its decimation by US subs
10/25/44 Rain squalls affect the battle off Samar
10/30/44 A squall allowed US sub Salmon to escape her pursuers after an attack on a small Japanese convoy
1/1/45 The Bodenplatte attack in Belgium by the Luftwaffe was delayed several days due to the weather

As you can see, the effect of weather usually isn't on the combat, it is rather in preventing the combat for ever occurring. Weather prevented contact or allowed someone to escape pursuit. When weather did effect combat, it was usually in convoy battles where bad weather scattered the convoy and allowed subs to pick off the stragglers created with interference from escorts. However, convoy battles are multi-day battles and the bad weather probably didn't persist over the whole time period. Similarly, some carrier battles like Coral Sea lasted several days. A number of the battles listed above are aircraft battles but, again, the effect of weather was usually to prevent, delay, or diminish the battle. Like convoy and carrier battles, air battles could cover a vast area and so have multiple weather conditions.

So, I'm not sure how I'd store the weather information since it could be different on different days and even different locations on the same day. Therefore, I think that I'll be satisfied with noting the influence of weather in the battle description (and animated battle map, if any) and then doing a text search like I did above. Thanks for the suggestion anyway; it is something that I hadn't thought of.
alecsandros
Senior Member
Posts: 4349
Joined: Wed Oct 14, 2009 2:33 pm
Location: Bucharest, Romania

Re: Advice wanted

Post by alecsandros »

Wel,, I'm a wargame enthusiast myself, so I couldn't help imagining things :) How long did it took to model all the battles and ships... ?

As for the weather effects - severe weahter can also reduce the probability of shell hits, as it degrades fire control. This is less of a problem for radar-equipped ships, and ships that have automatised turret and gun elevation incorporated (RPC). But for any other ships, firing on 6-meter waves is more difficult than doing the same on normal weather... (ex: battle of Samar, where Japanese ships were severely hindered by the rough seas and heavy rain - problems which appeared not to affect as much US DDs, which mantained accurate fire throughout the battle, thanks, in part, to radar and RPC)

I don't know how this could be incorporated in teh database though ?
Keith Enge
Member
Posts: 138
Joined: Sat Jan 01, 2011 1:36 am

Re: Advice wanted

Post by Keith Enge »

The WWII battle where weather had the most effect on gunfire was the 3/22/42 Second Battle of Sirte. The weather was so bad that Italian destroyers left early. This didn't help much, two capsized before they could reach safety. This battle obviously should have been in my list but I didn't put "storm" in my list of searched words. Adding that word, this battle and three more convoy battles were found. That isn't surprising; convoy battles lasted so long that most probably encountered bad weather sometime along the way.
How long did it took to model all the battles and ships... ?
It took quite a while. Fortunately, it was long ago. As I recall, entering the ship data took all of my spare time for four months or more. Remember, most ships underwent refits so have multiple data versions per ship. I think Renown has the record at 16 different versions. I only added planes and battles to the database about six years ago; those have been added periodically ever since (as well as a few more ships added or modified or getting additional versions). I added sinking locations and building shipyards for ships at the same; 8510 ships took awhile (I had previously done the fates). Adding animated battle maps are a somewhat lengthy process since the battles can be stepped through minute-by-minute. So I have to plot ship courses, gunfire (primary and secondary), torpedo spreads, and smoke screens. For carrier battles, air strikes, CAP, and searches must be featured. So, except for the simplest battles, making a map is an all day process even if I already have all of the needed info at hand.
Keith Enge
Member
Posts: 138
Joined: Sat Jan 01, 2011 1:36 am

Re: Advice wanted

Post by Keith Enge »

alecsandros -

I forgot to mention that the batle list created by your choosing of battle parameters could also detect the presence of specific individual ships. You were allowed to choose two ships and, when a list of battles was displayed, a visual flag was attached to the line for each battle that featured those ships, a specific flag for each ship (or both flags if both were present). However, this didn't limit to display to only battles with those ships; you had to look for the somewhat inconspicuous flags.

So, last night, I decided to let you search for battles that must have those ships. You have five options: ignore the two ships during the search, find battles with the first ship, with the second, with either, or with both. Since that was relatively painless, involving a minimum of programming since I made the program structure fairly easily extensible, I then did the same for specific planes. Instead of two ships, you have two groups of nine planes each. The same options apply: ignore, at least one plane in the first group, one in the second, one in either, or one in both. Combining these new ships and planes options, you can now find all battles where some Ju 78Ds attacked a British force including the carrier Formidable defended by either Fulmars or some version of Sea Hurricanes. I don't really know why I didn't implement this before; it now seems obviously useful.
alecsandros
Senior Member
Posts: 4349
Joined: Wed Oct 14, 2009 2:33 pm
Location: Bucharest, Romania

Re: Advice wanted

Post by alecsandros »

That's very nice, Keith.
I don't know if the program can select types of ships engaged. For instance, select all battles in which one force had at least 1 carrier, while the other had at least 1 BB?
Keith Enge
Member
Posts: 138
Joined: Sat Jan 01, 2011 1:36 am

Re: Advice wanted

Post by Keith Enge »

alecsandros -

Yes, that capability was already present; it's what I meant by "specific ship types from specific navies". The list follows:

07/09/40 Calabria
09/29/40 Operation MB 5
11/11/40 Taranto raid
11/27/40 Cape Teulada
03/28/41 Matapan
09/26/41 Operation Halberd
12/07/41 Pearl Harbor
03/08/42 "Tirpitz on the prowl"
03/22/42 Second Battle of Sirte
04/05/42 Indian Ocean raid
05/26/42 Prequel to Midway
06/04/42 Midway
06/14/42 Operations Harpoon and Vigorous
06/26/42 Russian convoy PQ.17 (although Tirpitz soon returned home)(and the carrier Victorious was only in the distant covering force which withdrew)
08/24/42 Eastern Solomons
10/26/42 Santa Cruz
04/03/44 Operation Tungsten (RN carrier attack on Tirpitz in fjord)
06/19/44 Philippine Sea
10/22/44 Battle for Letye (includes all four parts of the battle)
10/24/44 Sibuyan Sea part of Battle for Leyte
10/24/44 Part of Sibuyan Sea focusing on attack on Musashi
10/25/44 Cape Engano part of Battle for Leyte
04/07/44 Yamato's suicide run

A note explaining the "prequel to Midway" is needed. That "battle" is not a battle but rather a documentation of forces; there are a few of these in the database. This "battle" even has an animated map so you can see the positions of the various task forces involved in the Midway operation, including the diversion attack on Dutch Harbor in the Aleutians.

The US side had TF16, TF17, TF8 (Aleutians), and a fleet train and destroyers at French Frigate Reef.

The Japanese side was extensively fragmented:
1st Fleet, Main Body
1st Fleet, Carrier Attack Force
2nd Fleet, Strike Force
2nd Fleet, Escort Force
2nd Fleet, Occupation Support Force
Dutch Harbor Attack Force
Attu Occupation Force
Kiska Occupation Force
5th Fleet, Main Body
and I even added "miscellaneous" (two seaplane tenders with a destroyer)(the seaplane tenders were supposed to go to French Frigate Reef but found it occupied)

The battle list is interesting and, I hope, useful to you. It is interesting, even when they were present in the same battle, how few times carriers attacked battleships. I should note that the CV attack that torpedoed Bismarck is not in the database. That part of the hunting of her down is not covered. However, her battle with Hood and POW and her sinking by KGV and Rodney are both in the database and both have animated maps. You specified battleships and carriers so the search didn't look for CVLs or CVEs. Therefore, the battle of Samar, although in the database with an animated map, wasn't found because the carriers were CVEs not CVs. Similarly, Gneisenau and Scharnhorst sinking CV Glorious (again with a battle map), wasn't found because my database considers them battlecruisers, not battleships.

Keith
alecsandros
Senior Member
Posts: 4349
Joined: Wed Oct 14, 2009 2:33 pm
Location: Bucharest, Romania

Re: Advice wanted

Post by alecsandros »

Hello,
You obviously worked very hard for this :clap:

Could you tell me more about: "04/05/42 Indian Ocean raid". What carrier vs battleship attack was performed there ?


A small addition: there were 2 or 3 other RN carrier strikes over Tirpitz. I think they were called "Goodwood 1, 2 and 3"
Djoser
Senior Member
Posts: 383
Joined: Fri Feb 03, 2006 6:45 am
Location: Key West Florida USA

Re: Advice wanted

Post by Djoser »

Wow Keith, you have clearly busted your ass to produce this. Quite an impressive achievement...
Keith Enge
Member
Posts: 138
Joined: Sat Jan 01, 2011 1:36 am

Re: Advice wanted

Post by Keith Enge »

alecsandros -

The Indian Ocean raid was the Japanese attack into the Indian Ocean. I have an animated battle map that lasts 10 days, April 1 through April 10. Japanese surface forces made a series of attacks off the northwest coast of India. Japanese carrier planes made attacks on bases in Ceylon (Sri Lanka) and sunk many merchant ships and several warships including a carrier and two heavy cruisers. Somerville wanted a night action so kept the British Eastern Fleet out of harms way during the daytime. Therefore, his four battleships escaped attack. However, those battleships were among the 29 RN warships present so the battle database search included this battle even though the battleships were among the ships that didn't see combat.

By the way, my list of battles with a BB on one side and a CV on the other isn't complete. When selecting Axis countries, I just had it include Germany, Italy, and Japan. I forgot about the Vichy French which weren't Axis but did fight Allied forces. Therefore, 7/3/40 Operation Catapult and 9/23/40 Operation Menace, the attacks on Mers-El_Kabir (Oran) and Dakar, respectively, should have made the list. The RN carriers really didn't do too much but they were there.

About Tirpitz raids, there were several other attacks by carrier planes on Tirpitz. However, I only included the first one (Operation Tungsten) in the battle database so that's the only one that it found. The database also has an extensive hypertext discussion of naval and aviation topics (equipment, history, strategy, tactics, doctrine, etc). As part of its list of timelines, I included one that I called "The Teutonic bogieman" which documented the life of the Tirpitz. That timeline lists the following carrier attacks:

04/03/44 Operation Tungsten gets 14 hits but none major
04/24/44 Two days of bad weather aborts planned attack
05/12/44 Aborted by low clouds
05/28/44 Bad weather caused another abort
07/17/44 Smokescreens and flak foil an attack
08/20/44 Bad weather again
08/22/44 Two attacks broken up by flak
08/24/44 Two minor hits
08/29/44 No hits

Bomber Command made attacks before this list and, of course, later made three more attacks that finished her.

That hypertext discussion part of the database is quite large, almost 6 megabytes of text and graphics. There currently are 808 topics spanning 2162 pages with 6431 hyperlinks connecting them. Much of the stuff is interactive. For example, I mentioned the timelines above. There are 30 of them covering various campaigns and diplomacy, wartime conferences, and even the development of electronic countermeasures. They can be view individually or custom timelines can be created. You select the timelines you want and it builds a custom timeline, interleaving the entries. So, if for some reason, you wanted to know what was happening at the same time in Malta, Burma, and the Aleutians, you could select those timelines. You could even have battle entries from selected regions and/or the introduction of new aircraft from selected countries interleaved in too. Maps are interactive with clickable hotspots for battles and airfields. There are lots of other interactive goodies too. Over the years and certainly recently, I have probably spent more time adding to and enhancing the functionality of the hypertext than I have adding battles, ships, or planes.
Francis Marliere
Senior Member
Posts: 276
Joined: Mon Oct 18, 2010 3:55 pm

Re: Advice wanted

Post by Francis Marliere »

Keith, you database looks impressive. is it available from Internet ?

BTW, the name of the French naval base is, at least in French, Mers-el-Kebir, not Mers-el-Kabir (may be a typo ?).

Best,

Francis
Keith Enge
Member
Posts: 138
Joined: Sat Jan 01, 2011 1:36 am

Re: Advice wanted

Post by Keith Enge »

Francis -

My spelling of Mers-El-Kebir was indeed a misspelling in the post; it was correctly spelling in the database.

The database isn't available on the internet. However, I could email you a copy. I would like some feedback on its features, suggesting ways to improve them or even adding new ones. Built into the database is 341 pages (404K) of help text to aid in using some of its more powerful features. Again, I would like feedback on how adequate that help text is for someone that has never seen the database before. One thing nice about the database is that it has virtually no keyboard input. One place that the keyboard is used is in a "fuzzy" search for a particular ship by name when you aren't sure of the spelling or even the order of words in a multi-word name. You can type in your best guess and it produces a list of the closest matches from which you can choose. The other keyboard use is to name user-created custom ratings. The various bits of ship data can be used as variables rating formulas that combine them and constants via mathematical operators and even "If test" operators; all of this, however, is done with a "point-and-click" interface.

There is a potential problem in emailing you a copy. The database is quite large and some email providers have limits on attachment sizes. The database is about 8 megabytes. That includes about 3 megabytes of graphics for the hypertext discussion. I have provision for leaving out the graphics and having them be replaced by "placeholders" but that still leaves the database at over 5 megabytes. That can be compressed to around 3 megabytes but that may still be too large for the email providers. Some providers also don't like attachments that are executables (.exe extender); they want to protect their customers and fear that executables could be malicious software. I have been able to get around this in the past by just changing the .exe to something else like .jpg and then merely having the recipient change it back upon reception.

Let me know if you're interested,

Keith
Post Reply