Benchmarking, the Battle of Qabatiya Junction

Updated 27/11/17

This is an attempt to use the wargames benchmarking norms to analyse my BBWW2B Qabatiya Junction scenario. It will give us a good opportunity to look at how practical the benchmarks are to use. It is also interesting that this battle is in the benchmarking report's original database as "Jenin". I used the database to double check my own research which was helpful and lead to a number of changes. The remaining discrepancies are caused by the difference between the scope of the "Jenin" battle and my scenario.

The table below has the pertinent information for both my scenario and the original "Jenin" battle in the benchmarking database. The attacker is Israel the defender is Jordan.

The first table gives the numbers in both my scenario and the benchmarking database. Squads have been left out as that measure appears broken. Weapons have been omitted as that is based on the broken squad count

Item Scenario Attacker Scenario Defender Database Attacker Database Defender
Width 3 3 2 2
Troops 10000 8633 10900 6160
Mortars 30 24 50 38
AT weapons 72 24 71 25
Tanks 100 160 100 40
Artillery 42 36 36 24
AA Weapons 0 24 53 24
CAS sorties 55 0 12 0
Casualties     225 200
Tank Losses     18 15
Advance 13.7   5  
Duration 19 19 12 12

The width is in km. The original database gives the attacking frontage at the start of the operation. The battle subsequently expanded once the Israelis were through the initial Jordanian defences and when the 37th Brigade made its flanking attack. Thus the frontages are very conservative and will lead to higher force densities.

BBWW2 only represents line infantry as stands, many of the supporting arms are abstracted. This makes extracting troop strengths tricky. What I have done is given an infantry brigade a nominal strength of 5000 and a (fully) armoured brigade a strength of 2500. I have assumed that a battalion has one third these totals. This works lovely for the Israelis who have a strength of 10900 in the database and 10000 in the scenario. Ostensibly it works less well for the Jordanians who have 6160 in the database and 8633 in the scenario. It seems likely that the "Jenin" battle in the database does not include the Jordanian 40th Armoured Brigade's counter attack which would fit the database's numbers and also account for the shorter duration of the battle in the database (12 vs 19 hours)  plus the fact the database only give the Jordanians 40 tanks (one battalion) and my scenario gives them 160 (three battalions).

The number of AA weapons for the Israeli side is much lower in my scenario than from the database. However this is a moot point as the Jordanians have no air-support.

The Close Air Support (CAS) in my scenario is randomly arriving, so the CAS sortie rate is flexible, the average likely rate is given. BBWW2B is somewhat coy as to how many aircraft a model represents. For this analysis I have assumed the same ratio as tanks, 1:10. There is a big discrepancy between my scenario ad the database, but this is accounted for by the fact my scenario covers the Jordanian counter-attack and the original database does not. The Israelis sent many missions against the approaching 40th Armoured Brigade.

There are no losses for my scenario as yet as its not been played. Expect an update when it has.

The advance distance noted for my scenario is based on the need to travel from one corner of the table to the other. Although my scenario is longer its additional length is not proportional to the additional advance required over the historical battle. This hints that the victory conditions may be out of the Israeli's reach and it will be interesting to see the actual results.

The duration is the maximum possible length of my scenario. Its quite possible the battle will fizzle out before the time is up.

One small issue is what to do with light AFVs such as the M3 with 90mm AT gun. Can they reasonably be called tanks or are they AT weapons? If we were analysing a 1973 battle what would we do about the BMP?

The plausibility scores look like this:

Item Scenario Attacker Credibility Scenario Defender Credibility Database Attacker Database Defender
Force Ratio: men 1.16:1 Centrality        
Force Ratio: mortars 1.25:1 Centrality     1.32:1  
Force Ratio: artillery 3:1 Centrality     2.84:1  
Troop Density 3.33 Centrality 2.88 Plausible    
Mortar density 10 Centrality 8 Centrality 17 13
AT weapons  density 24 Plausible 8 Centrality 24 8
Tanks  density 33 Plausible 53   33 13
Artillery  density 14 Centrality 12 Centrality 12 8
AA Weapons  density 0   8 Centrality 18 8
CAS sorties  density 23 Plausible 0   8 0
Advance 13.7 Plausible     5  
Duration 19 Plausible 19 Plausible 12 12

On the whole my scenario looks very plausible. We have discussed the lack of Arab airpower and the corresponding lack of Israeli AA and it's historically accurate. The only other place where there is a red box is for the Jordanian armour. It could be argued that as the majority of Jordanian armour has a counter-attack role using a defender benchmark is misleading. As an attacker the 53 tanks per km would be plausible.

The 13.7km advance is plausible though not central so looks like its going to be a stretch.

Once I had sorted a credible way of representing troop strengths using the criteria was straight forward. It gave me pause for though around the Jordanian tank density but I think that has been adequately explained. The mobile nature of this battle illustrates how difficult it is to come up with a reasonable figure for frontage, as it varies throughout the battle. Although the numbers would have made no difference in this case its unclear how to treat light armour.

Update

I have now played the scenario once. My thanks to my opponent Keith for an excellent day's gaming. I carefully laid out al the casualties so I could photograph and count them at the end of the battle. Unfortunately in an over-zealous fit of tidying up they got packed away before I got the camera to them. So, the figures below are from memory. The Jordanians came out much worse than the Israelis

In the first part of the battle, before the counter attack the Jordanians lost: 13 infantry stands (including 2 Egyptian), 5 M47 stands, the Long Tom stand. A further two infantry stands ran away as did one tank. In the second part the Jordanians lost 5 M48

In the first part of the battle the Israelis lost 2 infantry stands and in the second part they lost a Sherman stand and and infantry stand.

I have calculated that to render a BBWW2B stand combat ineffective takes on average 50 casualties (an essay on the methodology coming soon). If we treat the artillery battery as an infantry stand for now, that's 700 casualties in the first period of the battle. The Israelis suffered 100 Infantry casualties in the first part.

The tank casualties do not make much difference to the overall personnel casualties. US WW2 experience: http://www.dupuyinstitute.org/blog/2016/08/26/u-s-tank-losses-and-crew-casualties-in-world-war-ii/ indicates on average one casualty per knocked out tank. UK experience gives 46% casualties when a tank was destroyed by another tank, this included crew becoming casualties after bailing out. The British study seems to be what we need for our overall casualty count. The M47 has 5 crew, giving 25 crew in the lost M47s. giving 12 casualties making 712 casualties in the initial part of the battle. M48s have a crew of 4 giving 9 extra casualties for a grand total of 721.

The Sherman has a crew of 4 so gives 2 casualties, and the Israelis lost a further infantry stand giving a final game total of 152.

In the historical database (which only covers the initial attack and not the counter-attack) the Israelis took 200 casualties (given the rounded figures I guess that the numbers are estimates) and the Jordanians 250. Our figure of 100 for the Israelis and 712 for the Jordanians do not closely match.

There are 10900 troops in the table top scenario for the Israelis, so that's 1.3% casualties. There are 8633 Jordanians so that's 8.3% casualties. Referring back to the benchmarking page the Israeli losses are in the central range. The Jordanian losses are not but are in the plausibility range. The exchange ratio is nearly in the centrality range and is definitely in the plausibility range.

I have yet to do the work on how much damage it takes to render an AFV stand combat ineffective. For now lets assume it's 40% casualties (i.e. 4 tanks per stand) same as the infantry. So the Jordanians lost 40/160 tanks (25%), 20 in the first part of the battle and 20 in the second part. The Israelis  lost 4 out of 90 (4.4%) tanks all in the second part of the battle.

Remember the second part of the battle is not accounted for in the Jenin database. Historically the Israelis lost 12 tanks and the Jordanians 15 in the first part of the battle. Once more the Israeli figure is too low and the Jordanian figure too high, though the Jordanian figure is close.

Overall both nation's tank loses put them in the central part of the range according to the benchmarking.

In conclusion I am fairly happy with the numbers. BBWW2B does not give outrageous results. The Jordanian infantry casualties are perhaps a tad too high but it really needs another replay or two to be sure.